LIBRARY
POVAl ONTARIO MIISFIIM
Digitized by the Internet Archive
in 2012 with funding from
Royal Ontario Museum
http://archive.org/details/stratigraphicexcOObull
wmm
H
191
iWMW
r AND ARCHAEOLOGY OCCASIONAL PAPER 9
WILLIAM R. BULLARD, JR.
Stratigraphic Excavations
at San Estevan,
Northern British Honduras
YAL ONTARIO MUSEUM
UNIVERSITY OF TORONTO
J?lir MlftiiiOT 'i° MUSEUM LIBRARIES
3 1761 04313 4519
mBwm
Wmm
Mi
Occasional Paper 9
ART AND ARCHAEOLOGY
ROYAL ONTARIO MUSEUM
UNIVERSITY OF TORONTO
liam r. bullard, jr. Stratigraphic Excavations
at San Estevan,
Northern British Honduras
william r. bullard, jr. is Assistant Director of the Peabody Museum
of Archaeology and Ethnology, Harvard University, and former Field Director
of the British Honduras archaeological expedition of the Royal Ontario Museum,
University of Toronto.
price: $3.50
© The Governors of the University of Toronto, 1965
PRINTED AT THE UNIVERSITY OF TORONTO PRESS
Contents
Preface, 5
Introduction
Geographical description of northern British Honduras, 7
Resume of archaeological and historical knowledge, 8
Aims of the San Estevan work, 12
Description of the San Estevan site, 13
Description of the excavations
Introduction, 14
Structure I, 75
Structure II, 26
Pottery
Introduction, 29
The Vasquez ceramic complex, 31
The Barklog ceramic complex, 35
The Trial Farm ceramic complex, 41
Chronological position of the ceramic complexes, 46
Stone and miscellaneous artifacts, 52
Objects of chipped flint, 53
Objects of obsidian, 55
Objects of ground stone, 55
Worked sherds, 56
Matting, 56
Summary and conclusions, 57
References, 60
Figures, 64
Plates, 86
Preface
The first field season of the Royal Ontario Museum's archaeological expedi-
tion to British Honduras has already been reported in Occasional Paper 8
under the title Late Classic Finds at Baking Pot, British Honduras. In
October, 1961, Hurricane Hattie struck the country. It was in the after-
math of this disaster that the second field season was begun in February,
1962. Arrangements were made to work at a ruin close to the village of
San Estevan, in the Orange Walk District of northern British Honduras.
The San Estevan excavations got underway in mid-March with a crew of
from eight to ten workmen and lasted until early June. The expeditionary
equipment then was placed in storage and the collection from San Estevan
exported for study. The analysis was performed at the Peabody Museum of
Archaeology and Ethnology, Harvard University, where large comparative
collections from the Maya area were available. The report was prepared
during the remainder of 1962 and part of 1963. The whole specimens and
type collections of sherds are deposited in the Royal Ontario Museum.
Smaller type collections are also deposited with the Harvie Foundation of
Calgary and the Peabody Museum of Harvard University.
As in the first season, the expedition received the generous financial
support of the Globe and Mail of Toronto and the Harvie Foundation of
Calgary. To Mr. Kenneth E. Kidd, Miss Lucile Hoskins, and Dr. Edward S.
Rogers I wish to express my thanks once more for their encouragement
and attention to all my requirements, and to Dr. A. D. Tushingham for his
general supervision of the work and his assumption of responsibility for
editing the manuscript.
The Government of British Honduras again cooperated fully with the
expedition. The Archaeological Commissioner, Mr. A. H. Anderson, m.b.e.,
acted as liaison between the expedition and the Government. Over and
above his official duties, he and Mrs. Anderson offered hospitality and many
kindnesses for which my wife and I are greatly indebted.
The Belize Estate and Produce Co., Ltd., through their general manager
Mr. A. M. Hore, enabled us to employ their capable representative in
Orange Walk, Mr. Gerald V. Smith, as our agent. Moreover, they made
available storage facilities for the expeditionary equipment after the San
Estevan excavations had closed. In the village of San Estevan, Mr. Adolfo
Vasquez gave us permission to excavate on his property and was instru-
mental in procuring housing and labour.
1 am grateful to Dr. J. O. Brew, Director of the Peabody Museum of
Harvard University, for making the facilities of that museum available for
study of the collection and preparation of the report. I also thank Mr.
Robert E. Smith and Dr. James C. Gifford for their advice in matters
ceramic. The final architectural drawings and the pottery drawings are the
work of Miss Avis Tulloch of Cambridge, Massachusetts.
The labour foreman at San Estevan was don Jacinto Cunil. He was
capably assisted by his son, Eufracio. As this paper goes to press, I have
learned of Jacinto's death in his milpa at Soccotz on May 1, 1964. Jacinto
was sixty-two years old. He began his long archaeological career many
years ago with Dr. Thomas Gann. Later he worked for J. E. S. Thompson,
W. R. and M. D. Coe, A. H. Anderson, Linton Satterthwaite and others.
He worked for me in Peten, Guatemala, in 1958 and 1960 as well as at
Baking Pot in 1961. Jacinto also has contributed knowledge of his native
Maya culture to ethnological studies. Many of Jacinto's qualities have been
described by his compadre, J. E. S. Thompson (see The Rise and Fall of
Maya Civilization [University of Oklahoma Press, 1954], pp. 261-3).
For my part, I owe a great debt, not only to Jacinto's skill as an excavator,
but more especially to his loyalty, honesty, and persistence. One could rely
upon Jacinto Cunil.
William R. Bullard, Jr.
Cambridge, Massachusetts,
January, 1964
Introduction
GEOGRAPHICAL DESCRIPTION OF NORTHERN
BRITISH HONDURAS
For the purposes of this report, northern British Honduras is considered to
be that part of the country lying north of approximately the latitude of
Belize city and the southern end of Hillbank Lagoon. Excluded is the strip
of territory along the Guatemalan frontier west of the longitude of Blue
Creek Village, which is enivronmentally and, I believe, culturally more
closely tied with the Peten of Guatemala. The principal drainages included
within northern British Honduras are those of the New River, the Northern
River, Freshwater Creek and much of the drainage of the Rio Hondo
(Fig. l;p. 64).
The land is low-lying, maximum elevations being less than 200 feet
above sea level. Topographically, it is flat to slightly rolling. There are
extensive areas of swampland. Such major rivers as the New and Hondo
have little gradient and are sluggish, although they are deep in places.
During dry years, salt water has been known to work up the New River as
far as Hillbank Lagoon, more than 75 miles from the sea. Rainfall decreases
as one moves northward in the country. In most of northern British
Honduras the annual fall is about 50 to 60 inches, most of it falling between
May and November.
The region is underlain by limestones of eocene and miocene origin.
Where these are close to the surface — and the bedrock is often very close
indeed under the surface — the land supports a deciduous seasonal forest
containing such trees as the mahogany, zapote, gombo-limbo, and various
palms. The forest composition is generally similar to that of central and
southern British Honduras and the Peten district of Guatemala, but because
of the drier climate the trees tend to be lower: 50 to 70 feet high as against
70 to 100 feet in the other areas. Leached sandy sediments cover the lime-
stone in much of the area northwest of Belize to the New River and north
as far as the latitude of Orange Walk, and also in a narrow strip running
just east of the Rio Hondo. These soils are responsible for great tracts of
pine savannah with which are interspersed comparatively restricted areas
of low broadleaf forest. The pine savannahs are unsuited for agriculture
and for this reason were not occupied by the ancient Maya.
The modern population of northern British Honduras north of the
town of Orange Walk is mainly of Maya Indian or Maya-Spanish origin,
with Spanish the most commonly spoken language. South of Orange Walk,
Negroes predominate and one most often hears "Creole," a dialect of
English. The Spanish-speaking area is strongly agricultural. Corn and
beans are the staple crops, as they were for the Maya of ancient times.
These are grown by the ancient milpa, or slash-and-burn, techniques which
have proved best adapted to the tropical lowland conditions. A recent
development is the great expansion of commercial sugar plantations, espe-
cially in the area between the New and Hondo Rivers. Sugar cane has been
planted over much land formerly used for milpa crops. The inhabitants
of several villages which previously were self-supporting now depend upon
wage work in the cane fields and must import their staple foods from
elsewhere. (For more complete information on the environment and
economy of British Honduras see especially Wright et al. 1959.)
RESUME OF ARCHAEOLOGICAL AND HISTORICAL
KNOWLEDGE OF NORTHERN BRITISH HONDURAS
For some forty years, from the 1890's to the 193CTs, Dr. Thomas Gann
and his agents combed northern British Honduras for Maya artifacts.
There are few large ruin mounds which do not show the scars of his
endeavours. The results, published in a long series of reports and popular
books, are the basis for the little which is known about the archaeology
of the region {see Thompson, 1939, Appendix C, for a complete listing).
For the modern researcher, the value of Gann's reports is mainly in the
illustrations of pottery vessels and other artifacts. Too seldom can we
glean from the text adequate information about the associations of the
objects, the nature of the sites in which they occurred, and even the loca-
tions of the sites. Since Gann's time, a very little archaeological data has
been recovered in the area, most of it salvage from mounds torn apart
for road fill and building material. No scientific excavations had been
undertaken until our work at San Estevan.
The materials collected by Gann and others indicated that northern
British Honduras had been occupied at least intermittently by Maya
Indians from some time before the time of Christ until the present. The
earliest settlement for which there is evidence was during the Formative
period. Subsequent remains relate to the Early and Late Classic periods,
the late Postclassic period and the Spanish Colonial period. Maya Indians
predominate in the population of this part of British Honduras today.
Representing the Formative period are several vessels illustrated by
Gann from Nohmul and the area between Nohmul and the mouths of
the New and Hondo Rivers (e.g., Gann 1939, PI. 10, 1-3). More recently
Haberland (1958) has illustrated two vessels, probably Late Formative,
from Luisville in the same area. He also reports the existence of a circular
platform structure with inset stairway, which is said to have Mamom
(Middle Formative) sherds in association.
The Holmul I, or Matzanel, pottery style seems well represented in
northern British Honduras. This style was introduced to the lowland Maya
area at the very end of the Formative period and overlaps into the Early
Classic period. Anderson obtained a good representative collection from
tombs in a large mound between Douglas and San Pablo (Anderson and
Cook, 1944). The mound in question was being demolished for road fill
at the time. It is believed to have been part of Gann's Nohmul. Gann also
illustrates Holmul I vessels from the Nohmul area as well as from Santa
8
Rita (Gann, 1918, PI. 13, b; 1939, PI. 10, 4). The Nohmul area, which
apparently includes a sizable stretch of country between the New and
Hondo Rivers, has also produced Early Classic and Late Classic pieces,
the former seemingly better represented than the latter (e.g., Gann, 1911,
PI. 18, a; L939, PI. 2, 7,2-5, PL 3, 2; PI. 4, 2, 4; PI. 7).
Traces of Postclassic occupation are not uncommon in northern
British Honduras. Sites of this period concentrate in the Chetumal Bay
area, along the New River and Freshwater Creek, and in the lower drain-
age of the Rio Hondo. The most widely distributed objects are human
effigy incensarios. Other outstanding remains are the famous wallpaintings
at Santa Rita and the caches of zoomorphic figurines at that and a few
other sites {see especially Gann, 1901, 1911, 1918, 1927; Salisbury, 1897).
In addition to Santa Rita, Postclassic finds have been reported from
Saltillo. Consejo, Benque Viejo (New River), San Andres, Douglas, Noh-
mul, Guinea Grass and Honey Camp. Most are surface finds of incensario
fragments. They are from the summits or bases of ruin mounds which
are probably of older date than the period of the incensarios themselves.
However, Postclassic stone architectural remains were certainly present at
Santa Rita and must exist at other sites. Sites of the same period are known
in adjacent Quintana Roo, the best described being Ichpaatun near
Ciudad Chetumal (Escalona Ramos, 1946; Sanders, 1960). The northern
British Honduras Postclassic material equates culturally with the late
Postclassic Tulum period of Quintana Roo (Sanders, 1960), the Mayapan
period of Yucatan (Pollock et al, 1962) and the period of Topoxte in
central Peten (Bullard, 1961). These are the artifacts which were in use
in this area at the time of the conquest by the Spanish, and some of the
incensario and effigy forms may in fact have persisted until well into the
Colonial period.
At the time of the Spanish conquest of Yucatan, northern British
Honduras was part of the important and populous native province of
Chetumal. Roys (1957, pp. 157-65) has assembled the available geo-
graphical knowledge concerning this province, and Chamberlain (1948)
has described the history of its conquest. Its capital lay in the stretch of
territory between Chetumal Bay and Lake Bacalar in Quintana Roo, just
north of British Honduras and close to the modern Ciudad Chetumal.
There were villages along the shore of the bay, along the New River and
probably along the Rio Hondo. This is precisely the area in which the late
Postclassic incensarios and other remains are most frequently found today.
Roys estimates the southern boundary of the province of Chetumal to
have been just south of Hillbank Lagoon, running to the coast between
Northern River and Belize. There were also villages on the Belize River
which had close contacts with the Chetumal area in early Colonial times
and may have been part of the pre-Spanish native province. But on the
other hand, the Belize River villages may not have been settled until the
conquest.
Francisco de Montejo and his lieutenant, Alonzo Davila, entered the
Chetumal Bay area in 1528 during Montejo's unsuccessful first attempt to
conquer Yucatan from the east. Davilla returned in 1531-32 and attempted
to establish a Spanish settlement at Chetumal, but was forced by the
hostility of the natives to withdraw. The province of Chetumal was
finally brought to its knees in 1544-45 by a ruthless war of attrition con-
ducted by Gaspar, Melchor, and Alonso Pacheco. The Pachecos founded
the town of Salamanca de Bacalar which was to remain thereafter the
administrative centre of the entire region. The Chetumal Indians rose
in the Great Maya revolt of 1546-47, but Salamanca did not fall and the
region was pacified without serious fighting. The cruel campaigns of the
Pachecos effectively destroyed the native province. Many areas were
depopulated, their inhabitants fleeing southward to remoter regions. Some
undoubtedly joined the independent Itza of Peten.
Not much is known about the subsequent history of northern British
Honduras and Chetumal Province. During the early colonial period the
region was under at least nominal control from Salamanca de Bacalar. An
early church document (Scholes et ai, 1936-38, Vol. 2, p. 63) gives us
the names of villages which had visita churches in 1582. Many were
probably within the borders of the modern British Honduras, but the
exact locations of nearly all are unknown today. Father Bartolome de
Fuensalida, in the account of his famous missionary entrada to Lake
Peten-Itza in 1618 has left a detailed description of the journey from
Salamanca de Bacalar to Tipu, which was on the upper Belize River and
the jumping off point for Lake Peten Itza (Lopez Cogolludo, 1867-68,
Vol. 2, Book 9). The route went by canoe from Lake Bacalar to the
mouth of the Rio Hondo, then along the coast to the mouth of the New
River which was then known by its Maya name "Dzuluinicob." Proceeding
up the New River, the travellers passed the Indian villages of Ppuncuy,
Zonail, Holaptin, and Lamayna or Lamanay, most of which are included
in the 1582 church list. Of these New River villages, only Lamayna, on
Hillbank Lagoon, can be accurately located today. From Hillbank Lagoon,
the route led overland, probably via Paslow's Falls, to the village of Luku
on the Belize River, thence upstream to Tipu. The Tipu Indians acted as
intermediaries between the Spanish and the independent Itza of Peten.
Tipu is on the 1582 church list and was nominally under Spanish control
from Bacalar, but the Indians tended to turn apostate during the long inter-
vals between visits by Spaniards. The exact location of Tipu and of the other
Belize River villages is unknown today, although Tipu was in existence as
late as 1697. From Fuensalida's description, it was probably on either the
Macal or Mopan branches of the Belize River in the general vicinity of the
modern town of El Cayo.
In the late 1630's the Indians in the southern part of the jurisdiction of
Salamanca de Bacalar rose against the Spanish. The New River villages
were apparently abandoned at this time, their inhabitants withdrawing into
the jungle or going southward to join the Indians of Tipu. The courageous
Father Fuensalida led a small unarmed party to try to persuade the Tipu
Indians to return to the fold, but was unsuccessful, and his party barely
escaped with their lives. We hear nothing more of the old Indian villages
of northern British Honduras. Tipu, on the upper Belize River, returned to
10
Spanish control in the 1690's at the time of the conquest of the Itza of
Peten (Villagutierre, 1933), but is fate thereafter is unknown.
Lamayna or Lamanay is the only one of the old New River villages
whose site has been identified (Roys, 1957, p. 163). It is Indian Church
at the northern end of Hillbank Lagoon. The masonry walls of the church
still stand, rising as high as 4 m. in places. The plan is that of a typical
early colonial T-shaped ramada church. The masonry portion consisted
of the chancel with sacristy or choir rooms opening to north and south. The
rest of the T-plan was formed by the nave, which extended to the west.
This was a thatched ramada of perishable materials, and thus no trace
of it remains above ground. The Rev. F. de P. Castells has published a
plan of the church ruin, but mistakenly thought it to have been of ancient
Maya origin ( 1904). Gann has also referred briefly to the church building
at Indian Church (1926, pp. 63-4). Two stela-like stones apparently were
set up next to the church wall. One remains; the other, said to have been
elaborately painted, has been removed. Just south of the church is a rect-
angular mound resembling in every way a low Maya substructure platform.
Another mound is located just north of the church, but unlike the mound
to the south, has been torn apart for its stone. There are no other ruined
buildings near the church except for the foundations of a brick building
erected during the nineteenth century. The houses of Lamanay were
probably of perishable materials and lacked raised substructures, like the
houses of the modern Maya, so that they have left no surface traces.
This is undoubtedly the church and town of Lamanay mentioned in
the 1582 church list, and which Fuensalida reports as burned in 1641
(Lopez Cogolludo, 1868, Vol. 2, p. 447). Indian Church is also the site
of a large mound group of greater antiquity than the sixteenth-century
Lamanay. This ruin begins about a quarter of a mile north of the church
ruin and stretches a considerable distance along the lagoon shore towards
the entrance to the New River. There are at least seven large pyramids
and an undetermined number of smaller mounds. At least one sculptured
stela, badly weathered and seemingly without hieroglyphs, is present. This
ruin probably dates mainly from the Classic period, but I have picked up
fragments of Postclassic incensarios from the surface. Several of the larger
mounds have been badly damaged on their summits by treasure seekers.
During the last century, Indian Church was the site of lumber camps and
a sugar mill. In the 1860's it was raided by Indians from Icaiche, Quintana
Roo, and British troops were stationed there for its protection. Indian
Church is now completely abandoned and overgrown.
British logwood cutters were settling the coastal regions of British
Honduras and working up the streams into the interior of the northern
part of the future colony by the late 1600's. None of the known existing
early records of British Honduras mention an indigenous Indian popula-
tion (Burdon, 1931). In 1718, only twenty-one years after their conquest
of the Itza, the Spanish are believed to have sent a military force from
Peten to block British expansion on the Upper Belize River. It is not known
whether Tipu existed at this time. Belize was attacked by the Spanish
11
in 1730 via the New River route from Bacalar, as well as from the sea. But
despite continuing Spanish opposition, the British logwood cutters were
well established by the middle of the eighteenth century along the New
and Hondo Rivers and southward to the Sibun River. There seem to have
been no Indian settlements on the New River at this time, and none of the
place names of the early Spanish colonial period have survived. The present
Maya and ladino population of northern British Honduras has moved in
during the past 120 years; the majority coming as refugees from the bitter
fighting and hard times caused by the War of the Castes and its aftermath
the Mexican territory of Quintana Roo.
AIMS OF THE SAN ESTEVAN WORK
Our original plan had been to initiate a project at Indian Church, the old
Lamanay, during 1962. But delays and upsets caused by the hurricane of
October, 1961, combined with the difficulty of access to the site, made it
advisable to seek more limited goals for the 1962 work. The ruins near
San Estevan seemed typical of the medium-sized ceremonial centre ruins
in northern British Honduras, and damage due to quarrying and other
recent activity was comparatively minor. Moreover, the ruins were easily
accessible, close to a good labour source, and generally so situated that
work could be begun and carried out with a minimum of overhead expense.
The San Estevan excavations were restricted to what could be accom-
plished in one season of work. Two buildings were tested in order to
determine the construction sequences and to obtain samples of pottery
and other artifacts associated with the different building periods. We hoped
that this data would bear on a number of archaeological problems. Since
San Estevan is located on the midst of the area in which Postclassic
incensario fragments and other remains have been found, is within the
limits of the native province of Chetumal, and is on the middle reaches of
the New River, long an important route of communication, we hoped
especially to recover data which would bear on the later periods of Maya
archaeology. Was the region abandoned at the end of the Classic period
like the regions further south, and then reoccupied after an interval of
time? Or was native settlement continuous from Classic times through
the Postclassic period? We thought we might find enough intrustive
sherds or other clues to aid in solving the still unresolved problems of
Yucaten-Peten chronological correlations during the late Classic and Post-
classic periods. At the least, we hoped to find out how the cultural sequence
in northern British Honduras corresponded with that of central British
Honduras and Peten on the one hand, and the Yucatan area on the other.
In so far as our tests were aimed at the Postclassic period, the findings
were negative. We recovered not a single Postclassic sherd from the San
Estevan ruin. Nor, to our surprise, did we find even an adequate sample
of the Late Classic period, the period which is usually most abundantly
represented in the ruins of central British Honduras and Peten. Instead, the
bulk of our material pertains to the Late Formative and Early Classic
periods.
12
Description of the San Estevan Site
The San Estevan ceremonial ruin is about two and a half km. south-south-
west of the village of San Estevan and about two km. southeast of the
nearest bend of the New River. The terrain in the region appears virtually
level, and there is nothing which can be described as a hill. Nevertheless,
there are broad gentle swells, and the ruin stands at a slightly higher eleva-
tion than the surrounding terrain on all sides except the west. To the east,
the land falls away gradually to an area of swamp. To the west, there is a
perceptible rise of land between the New and Hondo Rivers. From the
higher San Estevan ruins, one can make out one or two mounds of the
archaeological site which Dr. Gann called Nohmul, about 12 km. to the
northwest between San Pablo and Douglas, and the range of low hills
bordering the Mexican side of the Rio Hondo is visible on a clear day.
This is limestone country with comparatively thin topsoil. The area
around the ruin and the lower mounds themselves have been heavily
cultivated for milpa crops for many years, and today the orginal primary
forest of sapodilla, mahogany, palms, and other limestone-loving broadleaf
trees is almost entirely replaced by dense young second growth known as
uamil. Uamil is much more impenetrable and difficult to work through than
high forest, and our mapping was greatly facilitated by the clearing and
burning for cultivation of much of the north part of the site during our
working season (PL I and II).
The main part of the ceremonial ruin is shown in Figure 2. It covers
an area about 280 m. north and south by 130 m. east and west, with the
main axis oriented slightly east of north. Most of the nineteen mounds
included on the map are arranged around three plazas, Plaza B being at a
slightly higher elevation than Plaza C and the smaller Plaza A, where
our work was centred, being higher than Plaza B. No buildings stand
intact, and only the scantiest traces remain on the surface of a few mounds
to denote the underlying masonry construction. The highest single mound
is Structure XV, which rises more than 15 m. from its base and was
probably the principal temple of the ceremonial centre. Next highest is
Structure III, which in absolute elevation is only 2.5 m. lower than
Structure XV, although it rises from a higher plaza level. Structure XIX,
at the extreme northern end of the site, has a raised court partly surrounded
by subordinate platforms and is one of the largest mounds in terms of
total bulk. The long mounds, Structures V, VII, VIII, XIII, (PI. I), and
XVI, have traces of minor platforms on their flat summits and may have
supported superstructures of perishable material. Stairway projections are
apparent on Structures XIII and XVI. Structure XIV, actually two mounds
in the centre of Plaza C, was clearly a ball-court and conforms in over-all
dimensions to the other Classic period courts in the southern Maya
lowlands (PI. II). Conceivably, another ball-court was between Structures
IX and X.
Approximately 150 m. west of Structure II, at the south end of the
site, is another large mound accompanied by several smaller ones. Although
13
they form an integral part of the San Estevan ceremonial centre, we did not
include them on the map because they were in very dense growth and it was
impractical to cut survey lines and make the necessary clearing because of
different property ownership.
We found no certain evidence of stelae or other monuments at San
Estevan. However, the butt of an apparently upright stone, badly shattered
by milpa fires, is visible in the centre of the stairway projection of Struc-
ture XVI and may have been a stela. If other monuments once existed in
the plazas, it is probable that they would have been broken by the frequent
milpa fires and likely that the fragments would have been carried away
during the sporadic stone removal operations to which the site has been
subjected.
Quarrying of stone from the mounds for use as road fill and in modern
construction has taken place from time to time. Stone has been taken
from the south end of the ball-court mounds, from the northwest corner
of Structure VI, and from the sides of Structures IV, V, and VIII which
front on Plaza B. Fortunately, the quarrying has not been on a large enough
scale to seriously damage the interior structure of the mounds, but about
100 m. south of the ceremonial group on another landholding a complex
of small mounds has been almost entirely demolished to obtain stone.
There are two probable treasure-hunter's excavations at the site. One
is a trench which enters the top of the high Structure XV from the west
side and has almost completely destroyed the summit. The other is a pit
in the top of Structure IX. Conceivably, these are the work of Dr. Thomas
Gann who worked near San Estevan around the turn of the century
(Gann, 1895-97, p. 314; 1911, p. 86), but the older natives of the
village believe that Gann worked in mounds closer to the river.
To the east of Plaza C and of Structure XIX is a broad depression
where the limestone bedrock has been cut away. This was clearly an
ancient quarry for construction material. Scattered over the area around the
ceremonial centre and between it and the river are numerous low mounds
and mound groups which are presumably remains of domestic houses.
Some are two to three m. high, others barely break the ground surface.
Many are in little court arrangements of two or three or more. Although
such mounds are numerous in the vicinity of San Estevan, they do not form
a compact settlement comparable to a modern town. Instead, the cere-
monial centre seems to have served a rather dispersed population, as has
been reported elsewhere in British Honduras and Peten (Bullard, 1960;
Willey, Bullard, Glass, and Gifford, 1965).
Description of the Excavations
INTRODUCTION
We conducted excavations in two mounds, Structures I and II, which
border the south and west sides respectively of Plaza A at the extreme
southern end of the ceremonial centre. This part of the site was the most
14
feasible for excavation during 1962, since it was not about to be placed
under milpa cultivation as was most of the northern part. We started with
Structure I, which was of medium size compared with the other mounds
at the site. It was somewhat longer and broader in relation to its height
than either of the adjacent Structures II and III, and it was thought that the
structural remains would be less damaged by erosion than in the other
steeper mounds.
Most ancient Maya structures represent several building periods, often
spanning, in total, long periods of time. Not only were minor additions and
modifications frequent, but when a building was abandoned another might
be built over it completely concealing the earlier construction. A major
purpose of the work at San Estevan was to obtain a ceramic and artifact
sequence in good association with such an architectural stratigraphy. Since
we especially wanted information on the later occupation of the site, we
devoted much effort to the final building period and intended the deeper
excavations to be mainly stratigraphic tests. As the work on Structure I
progressed, sherds and other artifacts in good association with the latest
building period turned out to be comparatively scarce and the sherds badly
weathered. Culturally more informative deposits were found deep in the
mound interior, but these pertained to the earlier construction periods.
When the season was well along, I decided to excavate Structure II in order
to increase the sample from the late period and also to provide a check
on the stratigraphy encountered within Structure I.
Before describing the architectural features found, it is desirable to add
a word concerning the terminology used in this report. Ancient Maya
buildings consisted basically of two parts, a substructure and a super-
structure. The substructure is a raised platform with a solid fill and
masonry retaining walls. It supports on its summit the superstructure,
which is reached by means of a stairway. The superstructure may have one
or several rooms and walls of stone masonry or of perishable material.
Roofs were sometimes of masonry and corbel vaulted, of beams and mortar
and flat, or thatched. In this report, features encountered in the excavation
are lettered and numbered in the order in which they were found, although
they are described in the order in which they were built. For example,
Floor 1 was the first floor discovered in the excavation, although the latest
floor built. The different building periods within a mound are termed
constructions and are lettered. Thus, in Structure I, Construction I-A was
the first building period to be uncovered and was the latest built. Construc-
tion I-C represented the third building period discovered and, as far as we
know, was the earliest building period of Structure I. Should some future
excavation uncover a still earlier building period beneath Structure I, it
could be termed Construction I-D without requiring alteration of the
nomenclature.
STRUCTUREI (Figs. 3-8; Pis. iv-xn)
Before excavation, no wall lines or other structural features were visible
on the surface of Structure I. It was an oval mound with basal dimensions
15
a little more than 30 m. east and west and 20 m. north and south. The
rounded summit was about six m. above the ground level of Plaza A. The
slopes of the mound on the north, or plaza, side were more gradual than
on the south side, where there was a steep drop to the lower terrain lying
south of the ceremonial group.
After the mound had been cleared of all vegetation, except for a few
small trees left temporarily for shade, we trenched across the centre of the
summit at right angles to the long axis of the mound. The 10 to 20 cm.
of dark topsoil gave way to white powdery marl and squared blocks of soft
limestone. Soon we struck walls and the edge of a doorway. When the
trench reached floor level across the mound, we extended the excavation
and followed the floors and walls until the superstructure of Construction
I-A was completely cleared of debris. At the same time, we trenched into
the mound base on the north side, uncovering the edge of the latest plaza
floor and steps of the main stairway rising from the plaza. A stratigraphic
test pit was started through the plaza floor in order to determine the
sequence of floors and the nature of the plaza fill. Eventually, this test
reached a depth of about 2.5 m. below ground level, below the base of the
earliest construction period discovered in Structure I. Fill continues still
deeper. The surface of Plaza A is 3 m. higher than that of Plaza B
and apparently this rise is entirely artificial.
We completely cleared the main stairway, Stairway 1, and then tried
to follow the substructure base in a trench around the east end of the
mound. This part of the excavation proved the most difficult. The mound
slopes were very badly eroded, the upper parts of the substructure
retaining walls had completely fallen, and we even had trouble finding
the basal courses in some stretches. The late building fill of Structure I
had been composed of freshly quarried marl and stone blocks and contained
very few sherds or other cultural material, some parts of the mound being
completely sterile. The soft and crumbly limestone blocks used in the fill
and in masonry construction were little more than consolidated marl
themselves. The compacted fallen debris at the base of the substructure
was thus very similar in colour and consistency to the mound fill, from
which it was mainly derived. In places where the substructure wall could
not be found, we cut into the mound until we were certain we were in fill.
In Maya ruin mounds, at least the substructure base is normally preserved
by the mass of debris collapsed from above. That so little was preserved
here suggests that stone may have been stripped off after abandonment of
Structure I for re-use elsewhere. Hoping to find better preservation on the
other side of the mound, we attempted to locate the northwest substructure
corner, but met with even less success than at the northeast corner.
After the features of the latest building period, designated Construction
I-A, had been recorded, we started a trench 2 m. wide through the door-
ways of the superstructure, and a connecting trench 3 m. wide through
the centre of the stairway at plaza level. Floor 2, which we designated
Construction I-B, was found and cleared within the limits of the trench,
and then cut through. Below we came upon the well-preserved walls of
Construction I-C which were followed to their base on Floor 3. In order
to. follow out the plan of this interesting building, we widened the excava-
tion in the centre of the mound to about 8 in., necessitating removal of a
considerable amount of overburden. The substructure of Construction I-C
was followed to its base in a narrow cut along its west side. Finally, we
trenched under the floor of Construction I-C to below the level of the
substructure base, encountering no earlier building. At the close of the
work, the trench around the base of the mound and the deeper parts of the
excavations into the centre were backfilled.
TEST THROUGH PLAZA FLOOR
The test pit through the floor of Plaza A at the foot of the main stairway
of Construction I revealed a sequence of artificial fill deposits but did not
reach sterile soil (Fig. 3; PI. III). The earliest construction feature encoun-
tered was Plaza Floor VI which lies below the level of the base of Con-
struction I-C. The floor surface was of stucco or packed marl and, like
nearly all floors excavated at San Estevan, lacked a base of pebbles. The
floor dipped very slightly towards the south and disappeared before
reaching the southern edge of the test pit. The fill beneath Floor VI was
built up of irregular layers of marl and small stones interspersed with
lenses of dark soil. Sherd material was relatively abundant in this deposit.
Floor V was about 90 cm. higher than Floor VI. The fill separating
the two was similar in composition to that below Floor VI, but it contained
a much greater proportion of black refuse-laden soil than of marl. The
upper 20 cm. of the deposit consisted only of black soil and pebbles and
contained an especially large number of sherds. Floor V was a hard
plaster floor underlain by 15 cm. of nearly sterile marl. It reaches into the
mound as far as the foot of Stairway 2, which is part of Construction I-C,
and the two are surely contemporary. However, we could not clearly
determine a direct contact between Floor V and the stairway, and the
possibility exists that there was another floor surface between Floors V
and IV which we failed to identify in the field.
The various resurfacings of Plaza A represented by Floors IV through
I are plastered floors laid on fills composed of freshly quarried marl.
Virtually no sherds or other refuse material occured in these deposits.
Floors IV and III are associated with the original base of Stairway I,
Floors II and I with Stairway la which is an addition to Stairway I. The
surface of the final plaza floor, Floor I, had weathered away except for a
narrow strip abutting the stairway.
CONSTRUCTION i-c (Figs. 3-6; Pis. IV, V, VI, VII)
Construction I-C, representing the earliest building period of Structure I,
is composed of two more-or-less separable parts. The major part, which
was more fully investigated, is a small masonry-walled superstructure
building, probably a temple, standing on a substructure not quite 2 m.
high. The building opens on its north side onto a long platform oriented
with its long axis east and west so that in its entirety Construction I-C
17
had a "T" form. The east-west platform constitutes the second part of
Construction I-C. It had a stairway to the plaza on the north and it
apparently supported superstructure walls of lighter construction than
those of the main building. Since only a narrow strip of this platform
was exposed by excavation, and since it was not completely trenched
through, we do not know its complete plan nor the exact chronological
relationship between the two parts of Construction I-C. However, it is
probable that they were built at the same time and as a unit. We will
describe first the main superstructure building and its substructure, and
then the long platform on its north side. Various secondary modifications
and additions which were made before final abandonment of Construction
I-C will also be described.
The superstructure building has exterior dimensions of about 5 by 6 m.
The main entrance was to the north, facing the plaza. The upper part of
the walls had been sheared off when the later Construction I-B was
built over them. The back (south) wall stood intact to a maximum height
of 2.2 m. above floor level, the front wall to a height of 1.6 m. One course
of a projecting exterior molding was preserved at the top of the back
wall and is probably close to roof height.
The original floor, Floor 3b, was level throughout the building.
It was of stucco laid directly on the substructure fill. The walls were of
roughly squared blocks of limestone laid in courses. Mortar was apparently
of marl, and chinking stones were used abundantly. Wall width varied
from about 50 to about 80 cm. in different parts of the building. The
south and west walls had central recesses on the interior sides, and the
south (back) wall also had a projecting panel on the exterior. The interior
of the building was divided into two rooms, Rooms 1 and 2, by a segment
of wall and a masonry pier. Wall surfaces both on the exterior and
interior of the building had been well plastered and painted red (Munsell
Colour 7.5R 5/4 to 10R 6/4). Floor 3b and probably the later Floor 3a
has been similarly painted.
On the east side of the building, an interior stairway led upward from
Room 2 (Figs. 4, 5b). The entire northeast corner of the building had
been filled to support the stairway, and there was a pronounced outward
bulge of the exterior east wall at this corner. Treads and risers of the
interior stairway had been painted red. Since it is very unlikely that the
building had a second storey, the stairway must have led to the roof.
Windows, uncommon features of ancient Maya buildings, were present
in the south and east walls (Figs. 4, 5). In addition, there was a doorway
in the west wall which probably had been constructed originally as a
window. The windows and the doorway had had long wooden lintels set
into the masonry of the walls. The window in the south wall was in the
centre of the recessed area with its lintel 1.10 m. above floor level. As first
constructed, it was 45 cm. high by 85 cm. wide. Subsequently, the south
half of the opening had been twice reduced in size with stones and clay
so that, in its final form, the window was actually a wall niche with a
small opening to the outside (Fig. 5a). The doorway in the west wall had
18
its. lintel at the same height and was about the same width as the south
window. It is thought to have been originally built as a window because
the lintel is too low for a doorway without cutting away the floor beneath
and because the exterior Stairway 3 with which it connects is clearly a
secondary construction. The window in the east wall was smaller than
the south window but had not been reduced in size. The sill was nearly
flush with the tread of the second step of the interior stairway and its north
edge was flush with the riser of the third step.
A wall feature of unknown purpose occurs in both the south and east
walls of Room 2. The example in the south wall is more completely pre-
served (Fig. 5a). In the upper part of the wall and slightly to the east of
the centre of the recessed area is a wall niche which exposes a short
section of a long beam which had been set horizontally into the masonry
interior of the wall. The niche was 35 cm. high, 20 cm. wide and 30 cm.
deep. Its steeply sloping floor was coated with smooth hard plaster.
Plaster also filled the space between the exposed section of beam and the
back of the niche. Although the beam itself had completely rotted away, its
impression in the plaster was well preserved. It had been 10 to 15 cm.
in diameter and 4.2 m. long, running nearly the entire length of the
south wall. The example in the east wall (Fig. 5b) was at approximately
the same height and was similar in construction, except that the beam was
shorter and its north end terminated within the niche. Perhaps these
features were used to support heavy hangings of some sort, but the
presence of plaster between the beams and the back of the niches would
have prevented cords being wrapped around the beams.
Other wall features include a drain in the centre of the south wall,
running on a slight slope from floor level to the outside of the building
(Fig. 5a), and a small irregular cavity, in the east wall of Room 2, 60 cm.
above floor level (Fig. 5b). The latter might have held a projecting piece
of wood or stone.
Fallen sections of plaster in the debris filling the rooms provided valu-
able evidence concerning roof construction (Fig. 6a). Poles 5 to 10 cm.
in diameter had been set parallel and close together. Over them were placed
twilled mats which provided support for a plaster cap 15 cm. thick. The
upper surface of the plaster, forming the surface of the roof itself, had been
carefully smoothed and painted red like the rest of the building. Although
matting and wood had decayed, the plaster preserved the cast of the
matting and the contours of the wooden poles over which the matting
was laid (Fig. 6a). Very probably, the poles had been supported by
larger beams, but no evidence of these remained.
The substructure facing of Construction I-C was excavated only
along part of the west side of the building (Fig. 4; PI. vn). It was sloping
with apron and basal moldings and rose 1.9 m. above a plaza floor which
corresponded in level and was almost certainly continuous with Plaza
Floor V on the north side of Construction I-C. This floor did not run under
the substructure. Where protected by the fill of Stairway 3, a secondary
construction, the substructure facing was in excellent condition. It had been
19
covered with smooth plaster and painted red, with the exception of an
unpainted band about 40 cm. wide in the recess just above the basal
molding. The edges of the band were carelessly executed and drops of
paint had run down over the unpainted surface. The plaza floor abutting
the substructure base was also painted red. The fill of the substructure,
penetrated by a trench beneath the centre of Rooms 1 and 2, was com-
posed of lenses of pebbles, marl, and black refuse-laden soil. It may
have been laid up in sections, since a rough retaining wall was found. In
composition, the fill was similar to the fill beneath Plaza Floor V, and
these presumably contemporary deposits contained comparatively high
concentrations of sherds.
Modifications to Construction I-C include the following. The recess
in the south wall of Room 2 was walled up so that the window and other
features of that wall were concealed. The entire superstructure floor was
resurfaced and raised about 7 cm. (Floor 3a). The doorway in the west
wall was made at this time. It has been mentioned above that this doorway
probably was at first a window similar to the window in the south wall.
Because its wooden lintel was placed too low — 1.10 m. above floor level —
for convenient use as a doorway, a slot-like section of floor was cut
away beneath so that one stepped down 40 cm. before passing under the
lintel. On the exterior, Stairway 3 was built up against the substructure
to connect the doorway with plaza level.
As described in the beginning of this section, the superstructure building
opens onto a platform whose long axis runs east and west so that in its
entirety Construction I-C was more-or-less "T"-shaped. The length- of
the platform was not determined; its width was about 5 m. On its north
side, a main stairway (Stairway 2) led to plaza level. The steps were each
composed of two courses of limestone blocks and resembled in construc-
tion the steps of the superimposed Stairway 1 . The surface of the platform
is slightly lower than that of the main superstructure building, but it will
be noted on Figure 3 that both Floors 3b and 3a run without break over
the step. The fill behind the steps appears homogeneous with the fill
beneath Rooms 1 and 2, and it is likely that the two parts of Construc-
tion C were built together as a unit.
Near the south edge of the platform, running just north of the front
wall of Room 1, is a row of postholes which cut through the plaster floor
surface and are spaced 2.5 to 3.5 m. apart. Connecting with the line of
posts and in alignment with the north wall of Room 1 is a masonry wall
which is broken by doorways on each side (PL vii). The masonry wall
is clearly a secondary construction, but the posts were probably original
features of the platform. Probably postholes and walls exist on the other
sides of the platform but were missed by our trench. Presumably the
posts supported a roof of light construction, perhaps thatched. The sides at
first may have been open or walled with light materials such as poles
and plaster. Later, the masonry walls were added. In the west side of the
excavation, where we probed deeper, we found steps leading down against
the side of the apron molding of the substructure so as to connect the
doorway in the east-west platform with Stairway 3. The installation of
20
the masonry walls and these steps must have occurred at the same time as
the laying of Floor 3a, the closing of the recess in Room 2, and the
making of the doorway and construction of Stairway 3.
CONSTRUCTION 1-B (FigS. 3, 1\ PI. VIII )
No attempt was made to work out the plan of Construction I-B, and we
have knowledge only of those features which were directly superimposed
over the excavated part of Construction I-C. These include part of the
superstructure floor of Construction I-B, Floor 2b, and a minor secondary
modification represented by Floor 2a (Fig. 3).
Construction I-B represents a considerable enlargement of Structure
I-C. The roof and upper parts of the walls of the older Construction I-B
were torn out and the debris used to help fill the rooms. Varying materials
were used in the fill packed around Construction I-C. Part of the fill was
of freshly quarried marl devoid of cultural refuse. Other parts were of
stones, and there were occasional lenses of refuse containing ash, bits of
charcoal and comparatively large and well-preserved sherds. We found
sherds belonging to the same vessel in different parts of the fill, indicating
that the enlargement of the mound took place quite rapidly. A barrel-shaped
cache vessel with a lid (Fig. 16, a) was found in the fill of Construction
I-B near the northeast corner of Room 1 of Construction I-C and 50 cm.
below Floor 2b. It was designated Cache 2. This vessel was probably
placed in position when the later Construction I-A was built and was
inserted down through Floor 2b from above.
Floor 2b has three main levels ascending from north to south. In addi-
tion, there is apparently a step between the second and third levels.
The three levels probably correspond with three superstructure rooms, the
back rooms being higher than the front rooms as in the overlying Construc-
tion I-A. The third or highest level of Floor 2b had burned patches on
its surface. A secondary modification was the raising of the third level
(Floor 2a) and the enlargement of the step between the second and third
levels. The surface of Floor 2a was heavily burned and covered by a thin
layer of wood ash. We found no traces of masonry superstructure walls
within the excavated strip of Construction I-B. They may have been
missed by the excavation, superstructure walls may have been of perishable
materials, or wall stones may have been torn out upon abandonment
of Construction I-B.
The north edge of Floor 2b seems to be cut by the treads of Stairway
1, which is part of Construction I-A. This suggests that the stairway asso-
ciated with Floor 2b had been torn out upon abandonment. However,
the sequence of construction uncovered by our trench through the stairway
reveals nothing to support this interpretation. An alternate possibility is
that Floor 2b originally connected with the top of the sixth step of Stair-
way 1 and that the connection was broken when Construction I-A was built
and additional steps installed. If so, the original lower two steps of the
stairway which connect with Plaza Floors IV and III were probably
features of Construction I-B. Supporting this possibility is the fact that
21
these features directly overlie the stairway and plaza floor associated with
the earlier Construction I-C.
While searching for traces of the northwest corner of the substructure
of Construction I-A, we struck the buried corner of an earlier substructure
(Fig. 7). Because of the relative elevation of its base and top we suspect
that it belongs to Construction I-B but do not have sufficient information
to prove the association. The substructure facing slopes inward slightly
and is ornamented with apron and basal moldings similar to those of
Construction I-C. The surfaces are well plastered but without evidence
of paint.
construction i-A (Figs. 3, 8; Pis. IX, x, XI, xn)
Construction I-A represents the final building period of Structure I. It was
more completely excavated than Constructions I-B and I-C. Upon the
abandonment of Construction I-B, the substructure had been raised
about 70 cm. and a superstructure building with masonry walls erected
on top. The fill of the enlargement was composed of marl and blocks
of soft limestone. It was almost completely lacking in potsherds and
other refuse material. The overall east-west length of Construction I-A
was about 27.5 m., the width about 15 m., and the height of the floor
of the back room of the superstructure was 4.7 m. above the final plaza
floor.
Excepting the stairway, the substructure was in very poor condition.
The retaining walls were almost completely broken down, and only
with difficulty could we trace the substructure base around the east end of
the mound. The plan, as nearly as we could make out, is shown in Figure
8. It will be noted that the front room of the superstructure, Room 1, is
longer than the back room, Room 2. This difference is reflected in the
greater length of the front (north) part of the substructure over the back
part. The northeast corner has a series of insets. The northwest corner
surely had the same plan, but we were unable to follow the wall base there.
The most intact part of the substructure retaining wall was at the extreme
east end of the building where it had been covered and protected by the
marl fill of a low secondary platform. The wall stones had exposed faces
measuring about 40 by 20 cm. and were about 20 cm. deep. The wall rose
with a slight inward batter and had a basal molding rising 80 cm. above
plaza level. The maximum preserved height was 1.5 m., and nothing
remained to show whether an apron molding had been present or whether
the substructure had been of one or of two terraces. There were no sur-
viving traces of wall plaster. Our difficulties in tracing the substructure
retaining walls suggest that re-useable stone may have been stripped
off after final abandonment of Structure I. Stones were also missing
from parts of lower treads of the main stairway where collapse would
seem a less likely possibility than ancient stone-robbing.
The main stairway, Stairway 1, was 15 m. wide and ran across most
of the north side of the substructure. Each step was composed of two
22
courses of limestone blocks, the risers being 30 to 40 cm. high and the
treads about 50 cm. deep. The three protruding lower steps, designated
Stairway la, represent a secondary addition. Originally, the stairway des-
cended directly to plaza level, connecting with Plaza Floors IV and III.
It was mentioned previously that Stairway 1 may at first have risen only
as high as the top of the sixth riser of the original stairway so as to connect
with Floor 2b, and that it was part of Construction I-B. Thus, the upper-
most steps and Stairway la may represent additions made at the time
Construction I-A was built. The west halves of the fourth to eighth steps
(of the stairway in its final form) were set slightly forward of the east
halves, resulting in a pronounced curve or bulge in the centre portion of
these steps. The ninth step runs perfectly straight and may actually be part
of the building platform of the superstructure. A short step was placed
between the curving eighth and the straight ninth steps. From the tread of
the ninth step there is a further rise of 75 cm. to the level of Floor 1. This
is greater than the height of the other steps of the stairway. It is probable
that the arrangement at the head of the stairway was as restored in
Figure 3, but erosion has removed all traces except for the horizontal
location of the riser of the tenth step.
The superstructure building had masonry walls, a hard plaster floor
(Floor 1 ), and was divided into two long narrow rooms. The end walls of
both rooms were nearly entirely destroyed by erosion and the front and
back walls were also badly fallen. The maximum standing wall height was
1.4 m. on the median wall dividing the rooms. The room fill contained
very few potsherds or other artifacts. It was composed of fallen wall
stones and marl, the latter probably deriving from roof and wall plaster
and from crumbling of the soft limestone building blocks.
Walls were 1.0 to 1.1 m. thick and faced with rectangular blocks of
limestone laid in courses. There was no standard stone size, but an
average stone was about 20 cm. wide, 40 to 50 cm. long, and 20 to 30
cm. deep. The wall interiors, behind the facings, were of marl and small
unshaped stones. Wall surfaces had been covered with a plaster coating
2 to 3 cm. thick. The few patches of plaster which still adhered to the
walls bore no surviving traces of paint. The wall plaster curved out at
the wall bases to become continuous with the surface of Floor 1 .
Room 1, the front room, was 2.3 m. wide and 15.5 m. long. It was
entered from Stairway 1 by three doorways of which the central was the
widest. At the west end of the room, the floor terminated against a
plastered step about 30 cm. high. Since the walls at the ends of the
room had fallen, it was not possible to determine what the step repre-
sented. There may have been a raised bench at the end of the room, or
another smaller adjoining room somewhat similar to Room 3 of Structure
II. The east end of Room 1 had completely fallen away, but a similar
arrangement probably existed there. The median wall of the building
had been set slightly back from the edge of the platform forming the
higher floor of Room 2 so as to form a narrow ledge or offset in the
south wall of Room 1 .
23
Room 2 was smaller than Room 1, 13.5 m. long and 1.4 m. wide. Its
floor was 75 cm. above that of Room 1, from which it was entered by a
single doorway and a step which protruded into Room 1 . A ledge or offset
28 cm. wide and 40 cm. above floor level ran the length of the back wall.
There were burned patches in the central part of the floor in line with
the doorway, perhaps caused by the burning of copal incense or other
offerings. However, no pottery vessels or other artifacts were found,
and the floors of both rooms were notably clean.
Although no direct evidence was found concerning the construction
of the roof, it was probably of beam and mortar rather than vaulted. No
specialized vault stones were present, nor did the debris seem sufficient
for collapsed vaults. Moreover, the walls themselves, composed of soft
stone and marl, seemed structurally too weak to have carried the heavy
weight of masonry required by corbelled vaults.
Although the entire central portion of Construction I-A was removed
by excavation, we came across only two possible caches and one burial.
Burial 1. Cache 1 lay in the fill directly beneath the doorway of Room 2.
It consisted of small fragments of shell (ostrea) and sea-urchin tests.
They were scattered over an area about one square m. in extent. The
fragments, of marine origin, may have been thrown in the fill as a
purposeful deposit, but it is also conceivable that they are fossils deriving
from the limestone itself and that their presence here is fortuitous. Cache
2 was located below Floor 2b (part of Construction I-B) near the north-
east corner of the more deeply buried Construction I-C. It consists
merely of a barrel-shaped cache vessel with a lid (Fig. 16, a). No contents
survive. The vessel is nearly identical to one in Burial 1. It was probably
deposited when Construction I-A was built, its pit going deep enough
to penetrate Floor 2b.
Burial I was the only burial found in our excavations at San Estevan.
It lay in a stone cist at the bottom of a pit beneath the protruding step
between Rooms 1 and 2 (Fig. 3; Pis. XI, XII). The grave had been
dug down from the floor level of Room 1 before the floor had been laid
and the step installed. Thus it is contemporary with the building of Con-
struction I-A. The pit was roughly rectangular, 60 cm. wide by 2 m. long,
and was dug to a depth of 1.5 m., cutting through the floor of Construction
I-B into the fill of Construction I-C. The sides of the bottom of the pit
were lined with slabs set on edge. These were capped with two layers of
slabs placed horizontally so as to form a cist with an interior height of
40 cm. The extra-cranial skeleton was poorly preserved, and we could
only determine that the body had been placed extended, on its back, with
head to the east. Third molars were present and the sutures of the skull
were well closed, denoting that the burial was that of an adult. Further
observations were precluded by the condition of the bones. Several teeth
had jade insets. A small jade pendant (Fig. 18, /) was found under the
head and could originally have been at the neck or in the mouth. A basal-
flange bowl was inverted over the head (Fig. 16, </), two hemispherical bowls
were set on edge near the head (Fig. 16, b, c), and a barrel-shaped cache
vessel with a lid was placed at the left of the legs. The latter was badly
24
smashed by slumping of the lower tier of capstones of the cist. It was
nearly identical to the vessel designated Cache 2 (Fig. 16, a).
CERAMIC ASSOCIATIONS AND DATING
The dating of the building periods is determined by their association with
the ceramic complexes. These complexes are described in detail in the
pottery chapter of this report. From early to late they are:
( 1 ) Vasquez — Late Formative, probably contemporary with Chicanel.
(2) Barklog — early Early Classic, contemporary with Tzakol 1-2.
(3) Trial Farm — late Early Classic and Late Classic contemporary
with Tzakol 3 to Tepeu 2 (?).
Plaza Floor VI
Plaza Floor VI is the earliest evidence of finished construction found
in our excavations. The sherds in the fill below this floor were exclusively
of the Vasquez complex. No typological differences could be determined
between the pottery types below this floor and between it and Floor V.
Floor VI is probably contemporary with the Vasquez complex and,
consequently, Late Formative in age.
Construction I-C
Two cuts were made into the substructure fill of Construction I-C. One
was a trench beneath Rooms 1 and 2; the other penetrated behind Stairway
2. Since Plaza Floor V is associated with Construction I-C, the material
from below this floor is also of aid in dating the building. From all of these
deposits the sherds were overwhelmingly of the Vasquez complex. How-
ever, in the two cuts in the substructure fill there were a very few vessel
body sherds, both of the slipped and unslipped categories, which corres-
ponded in ware characteristics to types of the Barklog complex.
Construction I-C was buried while still in good condition. Thus, no
great length of time could have elapsed between its final abandonment and
its incorporation into the substructure of Construction I-B. The fill encasing
I-C included refuse lenses with well-preserved, apparently freshly broken
sherds of the Barklog complex. The number of earlier Vasquez sherds
mixed with them was negligible.
An interpretation based on the numerical predominance of Vasquez
sherds in the substructure would call for a Late Formative dating. But I
think a more reasonable explanation of the facts given above is that
Construction I-C was built at the beginning of the time of the Barklog
complex, that is, at the very beginning of the Early Classic period. Refuse
from the older Vasquez occupation of the site was used as construction
fill, but a very few contemporary Barklog sherds were mixed with it.
Construction I-B
The sherds from the fill of Construction I-B were almost all of Barklog com-
plex types, the few exceptions being of the earlier Vasquez complex. Con-
struction I-B appears to have been erected immediately after the abandon-
ment of Construction I-C, which is thought to be of very Early Classic date.
25
Construction I-B was succeeded by I-A which also is Early Classic,
although late in the period. Thus I-B's position in time is well bracketed
and it can be placed more or less in the middle of the Early Classic period.
Construction J- A
No form sherds were found in the almost sterile marl fill of Construction
I-A. The building date rests on the pottery associated with Burial 1, which
was placed in the mound when I-A was being completed, and also on sherds
found in the room debris and in other post-construction deposits. All of this
pottery is placed in the Trial Farm complex. The Trial Farm complex
probably lasted into the Late Classic period, but it must have begun in
the latter part of the Early Classic. The vessels from Burial 1 conform
most closely to Early Classic types, and one of the most ubiquitous forms
in the post-construction debris is the basal-flange bowl, a hallmark of the
Early Classic period. So, Construction I-A probably was built before the
end of that period.
In sum, the ceramic evidence indicates that the three building periods
of Structure I are all Early Classic (ca. a.d. 300 to a.d. 600 by the
Goodman-Thompson correlation of the Maya and Christian calendars).
Although the large number of Late Formative sherds in the deeper deposits
shows that there was a major occupation of the site at that time, the only
construction of the Late Formative period identified was Plaza Floor VI.
STRUCTURE II (Figs. 9, 10)
Structure II was a steep-sided pyramidal mound 7 m. high and about
20 m. across both the east-west and north-south axes. The only structural
elements visible on the surface before excavation were a few courses of the
exterior surface of the back wall of the superstructure (PI. XIII).
Excavation started at the summit. We immediately struck the inner
face of the back wall and followed it down to the floor level of Room 2 of
Construction II-A. The remainder of the intact portion of the superstruc-
ture was then cleared and, at the same time, we ran a trench 3 m. wide
up the east slope of the mound, exposing steps of the main stairway.
Erosion had destroyed much of the superstructure, and the stairway was in
much worse condition than the stairway of Structure I. Thus it seemed
not worth while to attempt further investigation of the substructure.
After recording Construction II-A, we laid out a trench 2 m. wide
in order to test the fill beneath the floor. The trench started at the base
of the west wall of Room 2 and extended 8 m. eastward through the
doorways of Rooms 2 and 1 . The excavation was carried downwards
until it struck Construction II-B, whose features within the confines of
the trench were cleared. We found that the east end of Construction II-B
had been torn out when II-A was built. At the west end of the trench, we
followed the substructure retaining wall of Construction II-B down to
about 2 m. from its top, which was as far as the limited working space
would allow. Excavation of this wall to its base would have required
considerable extension of the excavation and removal of up to 6 m. of
26
overburden. The final operation in Structure II was a test under the floor
of II-B in order to obtain a sherd sample, after which the deeper parts
of the trench were backfilled.
construction ii-b (Figs. 9, 10 b\ PI. xiv)
The limited exposure of the 2 m. wide test trench was insufficient to explain
the plan of Construction II-B or the intricate superstructure features
present (Fig. 10/?).
The substructure rose an estimated 3 m. above the original plaza
level. The fill, penetrated by a cut through Floor 3, was of dark gray
clay containing small stones and refuse. It produced a good sample
of well-preserved sherds. The substructure retaining wall on the back
(west) side of the building sloped inward slightly from bottom to top. At its
top was an offset and a further rise of 50 cm. to the level of Floor 3. In
the south half of the trench, the substructure retaining wall protruded
some 20 cm. further west than in the north half. If we assume that the
protruding section was part of a panel in the centre of the back side
of the building, then the centre line of Construction II-B must lie to the
south of the test trench and, therefore, south of the centre line of the super-
imposed Construction II-A. The retaining wall was of limestone blocks
measuring about 40 by 20 cm. on the exposed face and extending 40 to
50 cm. into the hearting. No evidence of plaster remained on this wall.
Floor 3, a plaster floor without a pebble base, has been twice resur-
faced. Along part of its west edge was a single course of stones about 15
cm. high, perhaps a footing for a wall of perishable material. Floor 4 is
30 cm. higher than Floor 3, with which it connects by a narrow doorway.
Rising from Floor 4 is a series of narrow walls of different heights. The
higher of these, at least, appear to have been sheared off when Construc-
tion II-A was built. Floor 5, of which only a small portion was excavated,
is 10 cm. lower than Floor 4 and connected with it by a step. To the
east of Floor 4, the construction had been torn away completely, probably
when Construction II-A was built. At the edge of the break in the south
side of the trench was a narrow remnant of floor which was at the same
elevation as Floor 3. The superstructure walls had been plastered, and
both walls and floors had been painted red. A deposit of ashes and refuse
covered Floor 3 to the north of the doorway connecting with Floor 4. A
similar deposit lay on the narrow terrace between Floor 3 and the edge
of the substructure retaining wall.
CONSTRUCTION II-A (FlgS. 9, 10 Cl\ Pis. XV, XVl)
The only excavated part of the substructure of Construction II-A was
a 3 m. wide strip up the stairway on the east side (Fig. 10 a). No steps
were complete across this strip, the lower ones being in particularly poor
condition, and it is possible that stones were removed after abandonment
for re-use elsewhere. Steps were of similar dimensions to those of Stairway
1 of Structure I. Between the third and fourth steps above plaza level
27
there seems to have been a platform or landing about 1.8 m. wide,
similar to but wider than the platform formed by the addition of Stairway
la to Stairway 1 of Structure I.
The substructure fill of Construction II-A was of marl and blocks of
limestone. Some stones had one plastered and painted surface and evi-
dently had been robbed from the superstructure of Construction II-B
which the fill encased.
Erosion of the mound slopes had cut away much of the superstructure.
Of the three rooms for which evidence remained, only Room 2 had a com-
pletely preserved floor area. The front room, Room 1, was 1.9 m. wide
and of undetermined length. It was entered from the stairway by a door-
way 3.4 m. wide. As in Construction I-A of Structure I, the floor of the
back room, Room 2, was 75 cm. higher than the floor of Room 1 and
was reached by a step which protruded into Room 1. Room 2 was 7.3 m.
long and 2.2 m. wide. Part of its back (west) wall was standing to a
height of 2.2 m. above floor level. A narrow ledge 15 cm. wide ran the
length of this wall 40 cm. above floor level. It was narrower but at the
same height as the similar ledge in the back wall of Construction I-A. A
doorway in the north wall of Room 2 led into Room 3, of which only the
end remained. Room 3 was the same width as Room 2, but its floor was
20 cm. higher. The step between the two was placed in the centre of the
doorway.
The room floors had been resurfaced. In Room 2 there were burned
patches in the centre of both the original floor and its later resurfacing.
Walls were similar in construction to those of Construction I-A. They
were slightly more than 1 m. in thickness, with the exception of the
partition between Rooms 2 and 3 which was 80 cm. thick. Wall surfaces
had been plastered. Small fragments of red-painted carved stucco orna-
mentation were scattered through the debris filling the rooms, but none still
adhered to the walls. The roof is believed to have been of beam and mortar.
CERAMIC ASSOCIATIONS AND DATING
Structure II had only two building periods, whereas in Structure I we
identified three. Nevertheless, the stratigraphy of the two mounds was
very similar and the results of each excavation served to check and
reinforce the other.
Construction II-B
The trench into the substructure fill of II-B produced an assemblage
of pottery types almost identical with that in the fill of Construction I-C
in Structure I. Vasquez sherds were by far the most numerous, but with
them were a few vessel body sherds of the Barklog complex. Immediately
after II-B was abandoned, lenses of refuse were dumped on its floor and
in the accumulating fill around the building. The sherds in this refuse were
apparently freshly broken fragments of Barklog vessels. These refuse
lenses and their contents closely resembled the refuse lenses in the fill
28
surrounding Construction I-C. There is little doubt that the earliest
buildings in both mounds were built at the same time and abandoned
at the same time. Consequently, Construction II-B can be placed with I-C
in the early part of the Early Classic period.
Construction II-A
Sherds in the substructure fill of Construction II-A included both Barklog
and Vasquez types, but none of the Trial Farm complex. The fill included
many limestone blocks which had been torn out of Construction II-B,
and this re-use of earlier building material probably accounts for most of
the Vasquez sherds. Sherds in post-construction deposits of II-A were
Trial Farm. II-A may have been built at about the same time as I-B, but
it evidently remained in use contemporary with I-A, which it resembled
architecturally. That II-A may have been older than I-A is suggested by
the resurfaced floor of the former; the floor of I-A had not been resurfaced
or obviously repaired during the lifetime of the building. In any event,
Construction II-A can be placed chronologically in the middle to later
part of the Early Classic period.
Pottery
INTRODUCTION
For the Maya area, the San Estevan pottery collection, comprising about
5000 sherds and 8 complete vessels, is a modest one considering the
amount of excavation accomplished. This is because large portions of the
fills of the excavated mounds were devoid of potsherds and because burials
and cache deposits were very few. Nevertheless, a clear-cut pottery
sequence was obtained from each mound. Three ceramic complexes were
determined by stratigraphic position within construction deposits. Each
complex was distinctive, with little overlap of types and little mixture of
earlier material with later in the deposits. The three complexes, with the
number of sherds used in their definition and their approximate chrono-
logical position, are as follows:
( 1 ) Vasquez complex ( 1330 sherds), Late Formative.
(2) Barklog complex ( 1350 sherds), Early Classic.
(3) Trial Farm complex (2130 sherds), late Early Classic and Late
Classic.
The Vasquez complex was named after the owner of the land on
which most of the San Estevan ceremonial site stands. Barklog and Trial
Farm are the names of adjacent properties.
The basis for classification of any collection should depend upon the
size, condition, and composition of the collection, as well as the amount
of prior knowledge concerning its relationships. The San Estevan pottery
collection, although obviously related to Maya pottery elsewhere, is in a
region where no detailed pottery analysis has previously been attempted
29
and where cultural differences may be expected. Since the collection is
comparatively small, the range of variation of many attributes is poorly
known; even many vessel shapes cannot be reconstructed in their entirety.
Thus, current methods of classification which stress minor variations in
attributes, and which are applicable to large collections, are not so useful
or feasible here. Adequate illustration and simple nomenclature have been
the goals.
The primary basis of classification is into "slipped" and "unslipped"
categories. In the Barklog complex the slipped category is divided into
polychrome and monochrome. In the Trial Farm complex difficulty was
experienced with even the slipped-unslipped breakdown because of the
high proportion of sherds whose surface had been completely weathered
away. Wares within the slipped and unslipped classes tend to be quite
uniform, although occasional individual sherds are ecountered which in a
larger, more representative, collection might be assigned to separate
sub-categories.
Within the major classes, breakdown is by vessel form, an attribute
which is one of the most significant for chronological studies of Maya
pottery. Minor variations in surface treatment, such as the presence and
absence of incision, are described under the particular vessel form to which
they apply. A good argument could be made for setting the Vasquez
complex red-and-cream dichrome apart from the monochrome of that
complex. This was not done because I found no form distinctions between
the monochrome and the dichrome; moreover, many weathered or fire-
clouded sherds might have belonged to either category.
Colour descriptions are according to the Munsell system (Munsell
Soil Color Charts, Baltimore). Commonly used shape and size categories
follow R. E. Smith (1955, p. 4) and J. E. S. Thompson (1939, pp. 68,
70). They are as follows. Diameters of vessels are measured from the
outside edge of the lip. Thickness of side is based on the average thickness
of the side below the rim: thin, 0.25-0.5 cm.; medium thick, 0.5-0.8 cm.;
thick 0.8-2 cm. Height of neck is taken vertically from the lip to the junc-
tion of the neck and side: low, 1-3 cm.; medium high, 3-6 cm.; high,
6 cm. and over. A plate is a vessel with an unrestricted orifice whose height
is less than one-fifth its diameter. A dish has an unrestricted orifice and
a height between one-third and one-fifth its diameter. A bowl has an un-
restricted or slightly restricted orifice, and its height may be equal to, but
not less than, one-fifth its diameter. A vase is a vessel whose height clearly
exceeds its diameter.
The paste of the San Estevan pottery has not been subjected to expert
microscopic examination. Thus, detailed descriptions of paste and temper
are not attempted. Generally speaking, San Estevan wares, both slipped
and unslipped, are tempered with either quartz or calcite. Tuff temper,
common in other parts of British Honduras, is either completely absent or
so scarce as to be negligible. Sherd temper may be present in some Vasquez
complex pastes, but cannot be confirmed without analysis. Where texture
description is attempted, the Wentworth scale (Shepard, 1956, p. 118) is
used as a basis.
30
THE VASQUEZ CERAMIC COMPLEX
PROVENIENCE AND GENERAL CHARACTERISTICS
The Vasquez Ceramic Complex dominates the deepest stratigraphic deposits
excavated at San Estevan. Specifically, these include ( 1 ) the fill deposits
beneath Plaza Floors V and VI, (2) the fill of Construction I-C (beneath
Floor 3b and behind Stairway 2), and (3) the fill of Construction II-B
(beneath Floor 3 of Structure II). The physical divisions between these
and later stratigraphic units were clear, as were the typological distinctions
between the Vasquez and the later ceramic complexes. Few Vasquez sherds
came from Barklog deposits; even fewer Barklog sherds were found in
Vasquez deposits. It did not prove possible to make chronological sub-
divisions within the Vasquez complex. For example, the same pottery
attributes were present in the deepest deposit tested, the fill beneath Plaza
Floor VI, as in stratigraphically later deposits.
Characteristic of the Vasquez complex is a red slipped monochrome
ware with a "waxy" feel. An important variant of this is a dichrome in
which one surface of the vessel, usually the interior, is slipped red and the
other surface slipped a cream colour. Simple incision is comparatively
common as a decorative device, painted designs being rare and restricted
to the most simple motifs. The characteristic vessel form, to which more
than one-half of the total form sherds belong, is an outcurving sided plate
with a flat bottom. Unslipped jars have very low necks, and, in contrast
with later ceramic complexes at the site, striation of jar bodies is numeri-
cally insignificant.
slipped ware (639 sherds)
The slipped pottery of the Vasquez complex is predominantly a red mono-
chrome with a gray to buff paste of fine to medium texture. Slipped surfaces
are well polished and lustrous with the "waxy" feel characteristic of much
pre-Classic ware in Peten. The slip is soft and tends to flake off, revealing
a cream to buff undersurface. Crazing is frequent. The most characteristic
surface colour is a darkish red (10 R 4.5/4 and near variants), with some
sherds of slightly more orange hue (in the 2.5 YR range). About 40 per
cent of the sherds show firing variations. Most of these are probably acci-
dental, since usually they affected only parts of the vessels and there are
intergradations between them. The most important are a darkening of the
slip through reddish-brown to black, and irregular mottling usually in red
and black. The mottling frequently follows craze lines. Less often, parts of
vessels have fired to a buff.
About 10 per cent of the slipped pottery is a red and cream dichrome.
The dichrome occurs on plates, dishes, and bowls, but no shapes were
found in the dichrome which were not also in the monochrome. The in-
teriors and rims are red, the exteriors have a cream slip (central colour is
10 YR 7/2) or, occasionally, are left unslipped. The cream slip often shades
to a buff, and often shows rootlet markings. Vessels with a cream slip all
over are very rare or absent at San Estevan.
31
The most characteristic decorative technique is a shallow pre-slip
groove incising made with a blunt tool. The width of the incisions is from
1 to 3 mm. They occur only as simple horizontal lines, often multiple,
encircling the vessel rims or exteriors. Sometimes they appear so casually
executed as to suggest that they were formed in the smoothing process.
Less common decorative techniques, all pre-slip, include groove incising
deeper than the above, chamfering of plate bases, finger impression, and
applique ridges. The last two seem associated only with the dichrome.
Painted decoration occurs rarely on the dichrome and consists of very
simple broad-line elements in red-on-cream. One or two poorly preserved
sherds may be decorated by negative resist technique.
Plates with Outcurving Sides ( 140 sherds; Fig. 11, a, 1-31 )
Form: Outcurving, rarely flaring, medium thick to thick sides; bottoms flat
or slightly convex when viewed from exterior. Rims are frequently thickened,
occasionally tapered, occasionally everted or downturned, but wide everted
rims are rare. Lips are rounded, pointed, occasionally flat or bevelled. Lip
diameters (based on 15 sherds): 18-34 cm., average 30 cm.; average
height: 4.2 cm.
Decoration : Monochrome vessels are slipped all over, including the bottom.
Approximately one-third of the sherds have very shallow pre-slip groove
incising. The incisions occur as horizontal lines, often multiple, on rim
interiors and upper surfaces and on vessel exteriors. Commonly, there are
one or two grooves on the exterior just above the basal angle (Fig. 11, a,
23, 26). The side just above the basal angle may also be chamfered or,
rarely, have a slight molding (Fig. 11, a, 24, 26, 27). The shallow groove
incision and the basal chamfering occur on both monochrome and dichrome
vessels.
Dichrome: Dichrome sherds comprise slightly more than 15 per cent of the
total sherds of this form. Ordinarily, interiors are slipped red, exteriors
have a cream slip or, more seldom, are left unslipped. Minor decorative
variants of the dichrome include: vertical stripes on the cream exteriors
(8 sherds; Fig. 11, a, 25-27); finger impressed exteriors (2 sherds; Fig.
11, a, 28); red and cream panels separated by raised ridges on exterior
(3 sherds; Fig. 11, a, 29, 30). One sherd of the last variant is the only
dichrome sherd found which had been entirely slipped in red before the
cream was applied. All others had the cream slip applied directly to the
paste. One weathered sherd, probably of this form, has a red exterior and
a cream interior with vertical bands seemingly done by resist negative
technique ( Fig. 1 1 , a, 31).
Dishes with Medial A ngles ( 6 sherds ; Fig. 11,/?)
Form: Medium thick sides recurving or sloping inwards. Base below
medial angle is convex. Of two rim sherds, one has a triangular bolster
rim, the other a plain rim with rounded lip. Lip diameters (2 sherds), 22
and 34 cm.
Decoration: Five sherds are monochrome, one is dichrome. Two mono-
chrome sherds have deep horizontal groove-incised lines on exterior (Fig.
32
11, /?, 1). The dichrome sherd has a red interior, rim and base. The
exterior is cream with simple broad-line criss-cross design elements in red
(Fig. 1 1 . /?, 3 ) . One other sherd may have had a cream base.
Hemispherical Bowls or Dishes ( 10 sherds; Fig. 11, c)
Bowls or dishes with rounded to nearly straight medium thick sides. No
base sherds identified. Rim slightly thickened on one sherd, plain on the
others. Lips pointed or rounded. Lip diameter (5 sherds) 18 to 34 cm.,
average 27 cm. Most sherds weathered but all apparently slipped red on
interior and exterior.
Bowls or Dishes with Vertical or Near-Vertical Sides and Thickened Rims
(8 sherds; Fig. 11, d)
Worm: Thin to medium thick vertical or slightly recurved sides. Rim
thickened or bolstered and sometimes slightly everted. Lip rounded. No
lower wall or base sherds were identified. Lip diameter (6 sherds) 26 to
29 cm.
Decoration: Two and possibly three sherds are dichrome with red interiors
and rims and cream exteriors; the rest are monochrome. One monochrome
has a rounded ridge on the exterior below the rim (Fig. 11, d, 5). One
dichrome sherd has a multiple groove-incised red band below rim on the
exterior (Fig. 1 1, d, 3).
Bowls with Slightly Restricted Orifice (6 sherds; Fig. 1 1, e)
Incurving or insloping medium thick sides; rim plain or slightly thickened;
lip pointed or rounded. No base sherds were identified. Lip diameter (4
sherds) 15 to 20 cm. Vessels were apparently slipped on both exterior and
interior. Three sherds have very shallow groove-incised horizontal lines
encircling exterior below lip.
Jars with Low Necks (6 sherds; Fig. 11,/)
Globular jars with low vertical or slightly flaring necks; angle at junction of
neck and shoulder is sharp; lip is pointed or rounded. Lip diameters (3
sherds) 10 to 19 cm. One sherd has slip on exterior of neck and upper
shoulder; the rest of the vessel was apparently unslipped. Other sherds
include only necks and were slipped on both surfaces. One neck has
shallow groove-incised bands on exterior.
Spout { 1 sherd; Fig. 12, a)
Spout tapers towards rim, is oval in section with lip diameters 3.0 and
1.9 cm.; length of fragment 6.3 cm. May have been placed vertically, but
no definite evidence. The exterior is slipped.
Sherds Possibly Later than the Vasquez Complex
The following four sherds come from levels which contain a predominance
of Vasquez sherds, but they differ in surface treatment and form from other
Vasquez types. Nor do they closely resemble types of the succeeding Bark-
log complex. The first three were found in the fill of Construction [-C. A
33
few Barklog sherds were identified in this stratum although Vasquez types
were by far the most numerous. The fourth sherd comes from a level
between Plaza Floors V and VI which contained only Vasquez types,
excluding only this large sherd which closely resembles a Late Classic type
elsewhere in British Honduras.
Brownware Cuspidor ? (1 sherd; Fig. 12, b, 1): Nearly vertical medium
thick sides or neck, slightly bolstered rim, flat lip, possibly globular body
below neck. Lip diameter 32 cm. Slipped inside and out with a glossy
reddish brown (about 3 YR 5/4) slip. The vessel surface below the slip
is not perfectly smoothed and, because of the resulting roughness, the
surface resembles the Barklog monochrome ware.
Z-angle Vessel ? (1 sherd; Fig. 12, b, 2): Vertical thin sides, protruding
basal or medial angle. Rim has slight labial flange and pointed lip. Lip
diameter about 14 cm. Sherd is weathered, but appears to have been
brownware like the preceding.
Flaring-sided Plate (restorable half vessel; Fig. 12, b, 3): Flaring medium
thick sides, flat base, lip diameter 12 cm. Interior is slipped orange (2.5
YR 6/8), rim is painted red (10 R 4/6), and the exterior below the rim
is unslipped. The ware and finish resemble Barklog pottery, but this shape
and decoration were not identified in that complex.
Brownware Vase (1 sherd; Fig. 12, b, 4): Nearly vertical medium thick
sides, lip diameter 19 cm. The interior and exterior have a brown slip
(5 YR 5/2) which is well smoothed but of low lustre. The exterior has a
simple decoration in parallel post-slip narrow incised lines. The paste is
pinkish with probably calcite temper. This sherd resembles and is identified
by Dr. J. C. GifTord as "Orange Walk Incised, Sotero Ceramic Group,"
which at Barton Ramie belongs in the early part of the Late Classic period.
Its presence in a deposit which otherwise contained only large quantities of
Vasquez (Formative period) sherds must result from an intrusion through
the plaza floor which we did not identify during excavation.
unslipped ware (688 sherds)
The unslipped ware of the Vasquez complex is well smoothed but not
polished. The paste is medium to coarse in texture. Surface colours range
from gray to buff. Most are in the colour range of Munsell 7.5 YR 6 to
7/2; dark extremes run to N 6/0, buff extremes to 2.5 YR 6/4. In marked
contrast to the later ceramic complexes, striated or raked ware forms an
insignificant proportion of the unslipped category. Of 571 unslipped body
sherds, only 31 (about 5 per cent) are striated, and most of these come
from a single provenience beneath the floor of Construction I-C and are
mostly from the same vessel. If we exclude this provenience unit, we have
a total of 468 unslipped body sherds of which only 1 1 (about 2 per cent)
are striated.
Plates with Flaring Sides ( 19 sherds; Fig. 12, c)
Flaring medium thick sides; rim usually slightly tapered; lip pointed or
34
rounded; base probably flat. Lip diameters (8 sherds) 16 to 30 cm.,
average 23 cm. A few sherds are weathered and might have been slipped.
Compared to the predominant slipped form, the sides seldom outcurve, the
rims usually taper instead of thickening, and there are no traces of groove-
Vessels with Widely Flaring Sides ( 14 sherds; Fig. 12, d)
Rim sherds of widely flaring or outcurving sided vessels. Side gradually
thickens towards rim, which is squared off; lip is pointed. No base sherds
identified. Lip diameters (11 sherds) 22 to 35 cm. These rim sherds are
probably from plates or dishes. Some could be from large storage jars with
widely flaring rims, but no other evidence for such vessels was found with
the Vasquez complex.
Bowls or Dishes with Rounded Sides (5 sherds; Fig. 12, e)
Medium thick rounded sides with plain rims and pointed or rounded lips.
One sherd from a bowl with slightly incurving sides has red pigment on
part of the unslipped surface as in the Daub Ware of Uaxactun.
Jars with Low Thick Necks (8 sherds; Fig. 12, /)
Low thick vertical jar necks with flat lips. Thick strap handles, whose
upper surfaces were flush with the jar lip, connected the neck to the
shoulder. Bodies were probably globular. Lip diameters (5 sherds) 12 to
18 cm. Neck height (6 sherds) 1.6 to 3.2 cm.
Jars with Very Low Necks or Collars (6 sherds; Fig. 12, g)
Medium thick sides, very low neck or collar which may flare slightly. No
base sherds identified. Lip diameter (2 sherds) 10 and 26 cm. Neck height
(4 sherds) 0.5 to 1.0 cm.
Miscellaneous Jar Sherds (3 sherds; Fig. 12, h)
(1) Neck-shoulder angle of a flaring necked jar (Fig. 12, h, 1); (2) jar
body sherd with raised finger-indented band (Fig. 12, h, 2); (3) flat base
of large jar (Fig. 12, h, 3).
Miscellaneous Bowl and Jar ? Sherds (2 sherds; Fig. 12, /)
( 1 ) Bowl with thickened rim and groove below rim on exterior, lip diameter
32 cm. (Fig. 12, /, 1); (2) rim of bowl or jar neck, slightly bolstered rim,
pointed lip, lip diameter 18 cm. Exterior has comparatively broad vertical
striations or raking (Fig. 12,/, 1).
Adorno (1 sherd; Fig. 12, /')
Fragment, oval in cross-section, with three appliqued pellets on one end.
THE BARKLOG CERAMIC COMPLEX
PROVENIENCE AND GENERAL CHARACTERISTICS
The Barklog complex comes from fill deposits laid down immediately after
the abandonment of Constructions I-C and II-B. In Structure I the deposits
formed part of the substructure fill of I-B, in Structure II of II-A. The fill
35
of each of these constructions was composed in large part of freshly quar-
ried marl or limestone blocks. These parts of the fill were nearly or entirely
devoid of artifacts. However, lenses and other limited deposits of ash and
refuse occurred which contained concentrations of large, well-preserved,
and evidently freshly-broken sherds. It is these sherds which form the basis
of the Barklog complex. In Structure I, the refuse lenses were in contact
with the walls of I-C both inside and outside of that building. In Structure
II, a refuse deposit lay on the floor of Construction II-B and another was
found adjacent to the west wall of its substructure. The separation between
the Vasquez and Barklog complexes was sharp. Comparatively few Vasquez
sherds were identified in association with Barklog sherds. These could be
easily segregated by shape and ware; moreover, most were more weathered
and battered than the Barklog sherds with which they were associated.
The Barklog complex is completely different from the Vasquez com-
plex in ware and vessel shapes. The "waxy" slipped ware of the latter is
replaced by a slightly thinner and often less smooth-surfaced ware which
is related to the "gloss" ware of central Peten. The sharp Z-angle bowl is
the characteristic vessel shape. Other common shapes include flaring sided
dishes and slipped jars with cylindrical necks. Flat basal supports are more
common than ring supports, although both occur. The elaborately painted
red, black, and orange polychrome of the Maya classic period appears in
combination with the basal-flange bowl, but both polychrome and the basal-
flange form are believed imported because of ware characteristics. In the
unslipped ware, striated jars, previously scarce, become dominant and
occur most commonly with high necks which are quite unlike the jar necks
of the Vasquez complex.
slipped ware: polychrome (16 sherds)
The polychrome pottery of the Barklog complex is decorated with red and
black designs on an orange ground. Munsell colours for the red range from
about 7.5 R 4/6 to 10 R 4/8. The orange is usually 5 YR 5/8 or 6/8
with the reddish extreme about 2.5 YR 5/8. One sherd has a red-and-black-
on-cream (7.5 YR 7/2) exterior decoration.
Surfaces are almost perfectly smooth and lustrous. The paste is distinc-
tive from that of the common slipped monochrome of the Barklog complex.
Of most sherds it is pinkish ( 10 R 5/4) in colour, is tempered with grayish
particles, and is of finer texture than the paste of the monochrome. The
paste of the two tetrapod vessels is buff (5 YR 7/3 and 10 YR 7/3) and
is tempered with very fine particles seemingly of calcite.
Basal-break Bowls with Tetrapod Supports (fragments of two vessels;
Fig. 12, k)
Form: Almost vertical straight, or slightly incurved, medium thick sides:
rounded or pointed lip; simple basal angle; convex base. Four large foot
supports were certainly mammiform in shape. Lip diameters 15 and 18 cm.
Height ( 1 vessel) 7.2 cm. Diameter of foot ( 1 vessel) 4.8 cm.
Decoration: All-over orange slip, excepting base of one vessel. Interior
36
decoration: red lip bordered with black band. Exterior decoration: panels
with curvilinear scroll elements alternating with plain panels which are
solid red or outlined with red.
Basal-flange Bowls (3 sherds; Fig. L3, a)
Form: Flaring, medium thick sides with flange at junction of side and base.
The one sherd which included the rim has a rounded lip and is from a
vessel with a lip diameter of 34 cm. Length of flange (2 sherds) 2.2 cm.
and 2.5 cm.
Decoration: Horizontal parallel lines and stripes. The most complete
specimen (Fig. 13, a, 1) has an exterior design composed of open work
crosses, parallel lines with dots, and curvilinear elements.
Rim Sherds, Probably from Basal-flange Bowls (7 sherds; Fig. 13, b)
Form: Flaring sided vessels with diameters about 30 cm. Rounded lips.
Most, if not all, of these sherds are probably from basal-flange bowls.
Decoration: Black (6 sherds) or red (1 sherd) lips. Interior: red or red
and black rim stripes. Exterior: red or red and black rim stripes below
which are rectilinear or curvilinear elements too fragmentary to identify.
One sherd has a red-and-black-on-cream design (Fig. 13, b, 4). All others
are red-and-black-on-orange.
slipped ware: monochrome (707 sherds)
The slipped monochrome pottery of the Barklog complex is nearly all red
ware. The most common colour is 9 R-10 R 5/6 and near variants; the red
extreme is 8 R 4/6, the orange extreme about 2.5 YR 5/8. Fire clouding
is present on a minority of sherds and causes colour variations through
reddish brown to black, less commonly to a brownish gray (10 YR 6/3).
True dense black ware is represented by less than five sherds of which only
two are form sherds. Black ware is not treated here as a separate category.
Slipped surfaces are lustrous and the ware is related to the Peten Gloss
Ware (Smith 1955, p. 27). However, apparently because of imperfect
smoothing before the vessels were slipped, the surfaces of most bowls and
dishes are slightly rough to the touch and the total effect is dull. On the
other hand, jar sherds tend to be more perfectly smooth and shiny. The
slip of the dull ware tends to flake off; that of the shiny ware is harder and
more durable.
The paste of the monochrome is gray to buff and of medium texture
with inclusions mainly of quartz. Paste of the jars and other vessels which
arc more highly polished tends to be more compact and somewhat finer in
texture with more variety in tempering material. But clear-cut differences
are not always apparent.
Sharp Z-angle Bowls (74 sherds and one restored half vessel; Fig. 1 3, c)
Vertical, rarely flaring, medium thick sides; protruding basal angle. Rim
has a slight bolster or is plain; lip is rounded or flattened, sometimes with
an interior bevel. Basal support is flat on most vessels, rarely a ring base.
37
Lip diameter (8 sherds) 24 to 40 cm., average 32 cm. Lip to basal angle
height (11 sherds) 3.3 to 5.7 cm., average 4.7 cm. The last measurement
excludes two exceptionally low sided vessels which have sides only 2 cm.
high. The red slip covers the interior and the exterior side including the
basal angle. The base was not slipped. One sherd, probably of this form,
has a glossy black slip.
Rounded Z-angle Bowls (19 sherds, from 2 or 3 vessels; Fig. 13, d)
Slightly flaring or outcurving thin sides. The protruding basal angle is
rounded off. Rim plain or bolstered, lip flat or rounded. Basal support flat.
Lip diameter 22 cm. Lip to basal angle height 3.3 to 3.8 cm. The red slip
is on the interior and exterior side including the basal angle. The base is
not slipped. This form differs from the preceding not only in the shape of
the basal angle but, especially, in the thinness of the side and base.
Flaring-sided Bowls or Dishes with Plain Rims (36 sherds; Fig. 13, e)
Flaring medium thick, rarely thin, sides; plain rim; lip usually flat, rarely
rounded. Lip diameter (7 sherds) 23 to 30 cm. The interiors are slipped,
the exteriors either slipped or unslipped. Two sherds are heavily fire
clouded and one may have an intentional back slip.
Bowls with Incurving Sides and Slightly Everted Rims (8 sherds; Fig. 13, /)
Bowls with slightly incurving thin to medium thick sides. Rim or collar
slightly everted. No base sherds were identified, but basal support probably
was flat. Lip diameter (4 sherds) 17 to 18 cm. Interior and exterior slipped,
except that base apparently unslipped on some vessels. Light pre-slip
incised lines may encircle interior of the rim.
Bowls or Dishes with Bolstered or Slightly Everted Rim (8 sherds;
Fig. 13, g)
Known only from comparatively small rim sherds and complete vessel form
or forms unknown. Flaring, occasionally vertical, sides; rim bolstered or
slightly everted and usually with an interior bevel. Lip diameter (3 sherds)
22 to 28 cm. Red slip on interior and usually on exterior.
Basal-flange Bowls (2 sherds; Fig. 13, h)
Protruding basal flange is at junction of base and side. One sherd has thin
walls and a ring base, the other has medium thick sides. Neither is a rim
sherd. The slip is on the interior and the exterior side to the tip of the
flange. The thicker sherd may actually be from a dichrome or polychrome
vessel. Both sherds have a smoother, more lustrous surface than the
majority of Barklog complex slipped sherds. The paste of one sherd is
pinker and finer textured than the usual Barklog monochrome and resembles
that of the polychrome. The other sherd has a more heterogeneous temper
than most Barklog sherds.
Basal-break Bowls with Everted Rims (2 sherds; Fig. 13, i)
Slightly flaring medium thick sides; thickened everted rim; rounded lip;
38
simple basal angle. Lip diameters 15 cm. One sherd slipped red, the other
a reddish brown (5 YR 4/3), which may be intentional or an unintentional
firing variation.
Covers for Bowls or Dishes ( 1 3 sherds, 1 2 of which are from the same
vessel; Fig. 13, /*)
Downward sloping from centre to rim. Rim is raised, its underside being
notched to engage the lip of the bowl. The most complete fragment had a
raised handle. Lip diameters (2 specimens) 23 and 25 cm. The upper
surface only is slipped. The larger fragment was slipped red and heavily
fire clouded. The other had a highly polished intentional black slip.
Jars with Medium-high Near-vertical Necks (131 sherds, representing 5 or
6 vessels; Fig. 14, a)
Globular jars with slightly insloping or slightly outflaring medium-high
necks. Sharp angle between neck and shoulder. Sides below neck are thin
to medium-thick. Jars with slightly insloping necks have rounded or pointed
lips; jars with flaring necks have a single lip groove. Basal support probably
ring (Fig. 14, a, 4) or flat, but no proven association of base and rim
sherds. Lip diameters (4 sherds) 9 to 12 cm. Neck height (5 sherds) 2.6
to 5.2 cm., average 4.0 cm. The exterior of the neck and body slipped red.
Slipped surfaces are smother and more lustrous than the surfaces of most
bowls and dishes of the Barklog complex.
Jar with Finger-indented Band on Shoulder ( 1 sherd; Fig. 14, b)
Sherd represents shoulder to junction of neck. Form of body and neck
unknown. Neck and upper part were slipped, rest of vessel apparently
unslipped. A raised finger-indented band is just below edge of the slip.
This is one of two Barklog sherds which have the red slip applied in
thin horizontal bands of differing density so as to render a streaky effect.
The same technique is characteristic of the type Sierra Red, Society Hall
Variety, at Barton Ramie and is there placed in the Formative period
Barton Creek Phase.
Potstand ( 1 partly-restored vessel; Fig. 14, c)
Vertical medium thick sides, flaring rim and base. Lip diameter 12 cm.,
height 6 cm. Openings had been cut through the centre of the sides while
the clay was damp, but the size and shape of these openings could not be
determined from the fragments. The potstand is actually dichrome rather
than monochrome. The lips of the rims and base were painted red; the
remainder was an unslipped but well-polished buff (7.5 YR 6/4).
Spout? (1 sherd; Fig. 14, d)
Cylindrical, diameter 2.5 cm. There are traces of a probably red slip.
Miscellaneous Bowl or Vase Rims (5 sherds; Fig. 14, e)
These probably represent different vessel shapes. All have vertical or near-
vertical sides, plain or thickened rims and lip diameters from 18 to 20 cm.
39
unslipped ware (580 sherds)
The unslipped pottery is gray to buff in colour, most sherds being around
5 YR 5/1. Dark extremes run to N 4.5, buff to 5 YR 6/4. Surfaces are well
smoothed but not polished. Paste ranges from medium to coarse in texture,
and the temper is apparently either quartz sand or calcite.
Slightly more than 80 per cent of the unslipped sherds are fragments of
striated jars. Of the remaining percentage a large proportion must be from
parts of the same vessels which did not bear the striations, such as bases
and necks.
Storage Jars with Striated Bodies (approx. 80 per cent of the unslipped
pottery; Fig. 14,/)
Form: Globular jars with medium-thick sides and flaring necks. One basal
fragment identified as from this form has a ring base. Necks are apparently
of two types.
(1) (6 sherds) high flaring necks with bolstered rim and rounded lip;
lip diameters (3 sherds) 24 to 26 cm. (Fig. 14, /, 1,2).
(2) (3 sherds) medium-high flaring necks with grooved lips; lip
diameters (2 sherds) 16 and 24 cm. (Fig. 14, /, 3).
Decoration: Body of vessels below neck striated with a multiple pointed
tool. The striations are commonly vertical or diagonal, occasionally criss-
cross. Neck type No. 2 has striations up as far as the rim. Other decorative
elements include appliqued discs on shoulder (3 sherds) (Fig. 14, /, 5)
and a finger-impressed band on shoulder (1 sherd) (Fig. 14, /, 6).
Miscellaneous Jar Necks (6 sherds; Fig. 14, g)
It is uncertain whether or not these were associated with striated bodies.
(1) (3 sherds) Outcurving or flaring necks, probably low to medium
high, with rounded lips (Fig. 14, g, 1 ).
(2) (2 sherds) Thick low widely-flaring necks with round lips (Fig.
14, g, 2).
(3) (1 sherd) Low thick vertical neck with flattened lip, similar to a
Vasquez complex form (Fig. 12, /), and perhaps derived from an earlier
deposit.
Incurved-sided Bowl (1 sherd)
Rounded incurving sides, plain rim with rounded lip.
Inner-inverted Foot Vessel (3 sherds; Fig. 14, h)
A flaring-sided vessel with open tubular projections rising from the interior
of the side. Rim is bolstered and lip grooved. This unusual form closely
resembles the dishes with four inner-inverted feet found in the Tzakol
phase at Uaxactun (R. E. Smith, 1955, fig. 17a). No definite evidence
has been uncovered concerning their use. Smith suggested that basket
cylinders may have been tied to the four tubular projections or "feef and
a cover placed on top so as to make a receptacle for seed which would let
in air and keep out rats {ibid., p. 127). Borhegyi (1959, pp. 54-5) thinks
40
them more likely to have been parts of pronged incensarios. Not improbably
they are related to a Teotihuacan form known as the "inside-handled
suspension bowl".
THE TRIAL FARM CERAMIC COMPLEX
PROVENIENCE AND GENERAL CHARACTERISTICS
The Trial Farm Complex represents the latest sherd material which could
be segregated stratigraphically at San Estevan. It comes from the debris
filling the rooms of the latest building period of both Structures I and II,
and from the debris and soil removed in clearing the substructure of Con-
struction I-A and the stairway of Construction II-A. The offerings found
with Burial 1 are also placed in this complex. Because much of the fill of
the latest building period was freshly quarried marl containing few or no
artifacts, it is probable that the majority of Trial Farm sherds are derived
from occupation debris of the latest constructions rather than from deposits
accumulated during the building-up of the mounds.
The outstanding characteristic of the Trial Farm sherds is the severe
weathering to which they have been subjected, a result of their exposure
comparatively close to the ground surface. Those sherds which are com-
paratively well-preserved do not appear different from Barklog sherds in
regard to ware characteristics, but vessel forms show marked changes. The
most prominent form, found in all provenience units, is the basal-flange
bowl. A few examples of this form were also found with the Barklog
complex, but were nearly all polychrome and were all believed imported
because of ware characteristics. In the Trial Farm complex, basal-flange
bowls completely replace the earlier Z-angle bowl form and are made in
the local ware. Other forms, including flaring-sided and incurving-sided
bowls and dishes and deeper bowl and vase forms, are specifically different
shapes from those found with the Barklog complex, although the small
sample sizes and very fragmentary nature of much of the material makes
knowledge of some complete forms uncertain. Ring basal supports become
dominant over flat supports. In contrast to changes in the slipped cate-
gories, unslipped storage jars appear comparatively unchanged.
Noteworthy is the apparent lack of mixture between the Trial Farm
and earlier complexes. Very few sherds of types specifically characteristic
of the Barklog and Vasquez complexes occurred in the surface debris.
For example, not a single sherd of the dominant Barklog form, the sharp
Z-angle bowl, was found in a Trial Farm provenience unit.
Because of the advanced degree of weathering, the Trial Farm complex
could not be divided reliably into slipped and unslipped categories as were
the earlier complexes. Of over 2100 sherds comprising the complex, only
about 17 per cent showed definite evidence of a slip and 18 per cent of
striations. The remaining 65 per cent were analyzed in a "weathered and
plain unslipped" category, the great majority of which were probably
slipped originally. Paste differences between slipped and unslipped were
41
not sufficiently marked to provide a criterion. For the purposes of descrip-
tion, use has been made of "probably slipped" and "probably unslipped"
headings.
Unweathered sherds of both the slipped and unslipped wares were
apparently similar to Barklog complex wares in colour, surface treatment,
and paste characteristics.
SLIPPED AND PROBABLY SLIPPED WARE
Polychrome and Dichrome (9 sherds; 1 complete vessel; Fig. 14, /, 16 g)
The severe weathering of most of the Trial Farm sherds probably is
responsible for the scarcity of identifiable polychrome and dichrome.
( 1 ) (1 sherd; Fig. 14, i, 1 ) Interior has a black rim stripe on an orange
slip. Exterior has a simple band design with dots in orange-and-black on
a cream (?) background. From a flaring-sided vessel with lip diameter
about 35 cm.
(2) (1 sherd) Black-on-orange exterior decorated with simple parallel
lines. Form unknown.
(3) (7 sherds, all from the same vessel; Fig. 14, /, 2) Unslipped vessel
of unknown form with traces of a simple broad-line design in either red or
black. The ware is not Slate ware.
(4) (1 complete vessel; Fig. 16, g) Flaring-sided, simple silhouette
dish with ring base. Lip diameter 43 cm., height 12 cm. Interior black
on red. Design, executed in a free-flowing, almost casual, style, may repre-
sent an animal or plant bordered with simple band elements. The vessel
form corresponds with that of the "Flaring-sided Dishes, Plates, or Bowls,"
Type 2 (see below), some of which probably also were dichrome or poly-
chrome. This particular vessel was found by accident in a house mound
about one half mile from the main ruin.
Basal-flange Bowls or Dishes (43 sherds, 1 complete vessel, 1 restorable
vessel; Figs. 14, /'; 15, a; 16, d)
Form: Flaring medium-thick sides, rounded or flattened lip, plain rims
except for one vessel which has bolstered rim. Probably ring base. Flange
at angle of base and side. Lip diameters 32 to 45 cm. Paste of most sherds
gray in colour and of coarser and looser texture than paste of the few
basal-flange bowls found in the Barklog complex.
Flange types: (1) (6 sherds; Fig. 14, /', 1) No sharp division between base
and flange. This type seems transitional between the sharp z-angle and the
basal flange. Compared to the latter, the sides are more flaring and the
angle between side and base more open. (2) (14 sherds; Fig. 14, /, 2-4)
Flange triangular in section, projecting 6 to 10 mm. from side of vessel.
(3) (12 sherds; Fig. 14, /, 5, 6) Flange a pointed or rounded ridge,
projecting not over 7 mm., and sometimes filleted onto vessel body.
(4) (9 sherds) Flange parallel sided or slightly triangular, projecting 12
to 25 mm. from side of vessel.
Decoration: Most sherds bear traces of a red slip. No certain evidence
exists that any were polychrome.
One restorable vessel (Fig. 15, a) has a piano-relief decoration on the
42
exterior side. The design includes raised rectangular interlocking elements.
The background of the design panel is textured by raking with a multiple
pointed tool. This vessel, which is badly weathered, probably had a black
slip except for the cut-out portions of the design.
Flaring-sided Dishes, Plates or Bowls (72 sherds; Fig. 15, b)
These are rim sherds only, no connecting base sherds being identified. Most
are badly weathered, but all or most were probably slipped and some
may have been polychrome or dichrome.
( 1 ) (46 sherds; Fig. 15, /?, 1-3) Thin to medium-thick sides, plain rim,
flat lip, lip diameters greater than 30 cm. Most of these rims are probably
from basal-Mange bowls or dishes. 9 sherds are slipped red on both sides;
2 sherds apparently slipped on interior only; remainder badly weathered.
(2) (15 sherds; Fig. 15, b, 4-6) Medium-thick to thick sides; pointed,
rounded, or occasionally slightly bevelled lip. Diameters greater than 35
cm. A few sherds probably are from basal-flange bowls, but the majority
are from simple silhouette flaring-sided dishes similar to the dichrome
vessel in Figure 16, g. All were probably slipped and two may have been
dichrome or polychrome.
(3) (4 sherds, probably from the same vessel; Fig. 15, b, 7) Widely
flaring-sided dish or plate with wide interior rim bevel. Apparently slipped
on interior only.
(4) (3 sherds; Fig. 15, 6, 8) Bowls with thin sides, diameters about
30 cm. The sherds are so badly weathered that no trace of slip is visible.
(5) (4 sherds; Fig. 15, b, 9, 10) Medium-thick to thick, slightly
outcurving sides, with plain or tapered rim. Lip diameters greater than
30 cm. Trace of red slip on one sherd. The other three sherds are dis-
tinguished by a pinkish paste. Possibly jar rims, but diameters seem too
great.
Dish with Lateral Flange ( 1 vessel; Fig. 16, e)
Widely flaring sides and slightly rounded base. The side-base division is
marked on the interior only, and the form closely approaches a simple
silhouette. A slightly projecting flange is on the exterior side about half
way between the rim and the base. Base has a ring support. Lip diameter
38 cm. Height 9 cm.
This vessel, the only example of its form found at San Estevan, comes
not from the main ruin but was salvaged from a house ruin about half a
mile from the main ruin. It was associated with the dichrome vessel in
Figure 16, g.
Bowls or Dishes with Incurving Sides (28 sherds; Fig. 15, c)
Vessels with incurving sides and slightly restricted orifices. Most sherds
have traces of a slip.
(1) (13 sherds; Fig. 15, c, 1, 2) Medium-thick sides, rim slightly
thickened, lip pointed or rounded. Lip diameters (6 sherds) 21 to 3 1 cm.
Rims with pointed lips tend to have sharper curve of side than do rims
with rounded lips, and some of the latter may actually be from barrel-
shaped vessels. About half of the sherds have traces of a red slip.
43
(2) (12 sherds; Fig. 15, c, 3-5) Thin to medium thick sides. Rim
usually thickened on interior and overhanging. Lip has a steep interior
bevel. Lip diameters 20 to 32 cm. (4 sherds) . Most sherds had a slip.
(3) (3 sherds; Fig. 15, c, 6, 7) Medium-thick to thick sides, rim
bolstered (1 sherd) or thickened with exterior ridge (2 sherds). Lip dia-
meter about 35 cm. One sherd has traces of a red slip.
Bowls with Slightly Incurving Sides and Everted Rims (3 sherds)
Apparently the same as a Barklog complex form and possibly derived
from earlier deposits (See Fig. 13, /).
Bowls or Dishes with Bolstered or Thickened Rims (8 sherds; Fig. 15, d)
A miscellaneous group sharing medium-thick to thick sides, bolstered or
thickened rims, and outcurving or slightly rounded sides. Lip diameters
26 to 45 cm. All seem to have been slipped red.
Bowls with Rounded Sides and Plain Rims (12 sherds, 2 complete vessels;
Figs. 15, e\ 16, b, c)
Thin, occasionally medium-thick, rounded sides. Rim plain, sometimes
very slightly thickened. Lip rounded, flat, or pointed. Bases with flat or
ring support. Lip diameters 17 to 24 cm. The two complete vessels from
Burial 1 (Fig. 16, b, c) have a heavily smoke-darkened red slip on the
interior. One is slipped on the upper part of the exterior; the other has
an unslipped striated exterior. About half of the sherds have traces of a
red or black (?) slip.
Cylindrical or Slightly Barrel-shaped Bowls or Vases (15 sherds, 2 re-
storable vessels; Figs. 15, /; 16 a)
Thin to medium-thick vertical or very slightly insloping sides. Rim
tapered or plain, with pointed or rounded lip. One vessel has a slightly
bolstered rim with flat lip. Base flat. Lip diameter (7 sherds, 2 vessels)
14 to 19 cm. About one half of the total number of sherds have traces
of a red or black slip. The two complete vessels have a heavily-smudged
red slip on the exterior. One body sherd has a basal molding, another
an encircling ridge.
The two complete vessels, one from Cache 2 and the other from Burial
1, had lids in place when found (see below Covers or Outcurving-sided
Plates).
Covers or Outcurving-sided Plates (2 vessels; Fig. 16, a)
Both vessels would have been interpreted as plates with widely outcurving
medium-thick sides and flat bottoms if they had not been found in place
as lids for barrel-shaped vases (see above). Lip diameters 17 to 18 cm.;
heights 3 to 4 cm. Both vessels formerly slipped red on interior and
possibly on exterior, but both are heavily smudged by burning.
Barrel-shaped Vases or Bowls? (4 sherds; Figs. 15, g; 16, /)
Incurving thin sides, pointed lip. Lip diameter between about 10 and 17
cm. Vessel form probably a barrel-shaped vase or neck-less jar. Sherds
were all probably slipped.
44
Basal Supports ( 1 I sherds)
Basal supports for which the form of the remainder of the vessel is un-
known include 9 ring supports, I flat base, 1 hollow foot (Fig. 15, h) .
Four ring supports are from vessels with red slips; the other sherds are
too weathered to determine whether slipped or not.
Slipped Jar Necks (5 sherds; Fig. 15, i)
Medium high, slightly flaring or vertical necks with plain rims. Lip dia-
meters 10 to 15 cm. Two sherds slipped red, others weathered but
probably originally slipped.
Carved Ware ( 1 sherd)
A fragment of a bowl or dish with a carved relief decoration on the
exterior. The sherd is too small and weathered for determination of the
nature of the design. The paste is calcite tempered.
Effigy Whistle (1 sherd)
A badly weathered and battered fragment of a figurine whistle. It probably
represented a bird.
UNSLIPPED AND PROBABLY UNSLIPPED WARE
JJnslipped Jars
Because of the advanced degree of weathering of most Trial Farm sherds
it is impossible to determine accurately the proportion of striated jar sherds
in the unslipped category. Only 341 sherds were identified as striated,
while 1158 sherds were classed during analysis as "weathered and un-
slipped plain/' But probably many weathered sherds were striated and
the proportion therefore substantially higher than the above figures seem
to indicate.
Striated jar sherds resemble those of the Barklog complex, except
that a large proportion of the well preserved sherds have traces of a reddish
wash lightly applied over the striations. Neck forms are as follows:
(1) (12 sherds; Fig. 15, /', 1-3) Outcurving necks with bolstered or
everted rims, medium-thick to thick sides, lip diameters 19 to 24 cm.
Most sherds probably from striated jars with medium or high necks.
(2) (5 sherds; Fig. 15, /', 4, 5) Outcurving necks with overhanging
or triangular rims and sharply pointed lips. Lip diameters 17 to 24 cm.
One sherd with overhanging rim has horizontal impressed lines on
exterior below rim.
(3) (3 sherds; Fig. 15, /', 6) Slightly Oaring or outcurving medium-
high necks with plain rims. One sherd has grooved lip; another has lip
with interior bevel. Lip diameter 12 to 18 cm.
Strap Handle ( 1 sherd; Fig. 15, k)
A jar handle attached either horizontally or vertically to body. Width of
handle, 3.5 cm. Has traces of a lightly applied red wash similar to that
on striated body sherds.
45
Bowls(?) with Triangular Rim Bolster (2 sherds; Fig. 15, /)
Medium thick, slightly rounded sides, triangular rim bolster, pointed lip.
Lip diameter more than 35 cm. Both sherds weathered, but originally
probably unslipped.
Bowls or Jars with Spikes and Piecrust Rims (10 sherds; Fig. 15, m)
Sides insloping or slightly flared. Finger-indented fillet on exterior of rim
with top of fillet nearly flush with lip. Spikes and possibly other ornamenta-
tion were appliqued to exterior. Spikes protrude from a minimum of 3
mm. to a maximum of 15 mm. Lip diameter 15 to 18 cm. These vessels
were quite likely incensarios.
CHRONOLOGICAL POSITION OF THE CERAMIC
COMPLEXES
SOURCES OF COMPARISON
The dating of the San Estevan pottery complexes, and consequently of
the building periods, relies upon comparisons with documented pottery
sequences elsewhere in the Maya lowlands. Most of the data appropriate
for comparison and now available comes from work within a few restricted
regions, particularly west central British Honduras and central Peten.
For large parts of the lowland area we have no pottery studies at all.
In northern British Honduras, no stratigraphic work has been done
previous to the tests reported herein. Many pottery vessels from the
region exist in private and museum collections (see Thompson, 1939,
Appendix c), but for the great majority we know next to nothing con-
cerning associations or exact provenience. The nearest useful strati-
graphic work is at San Jose (Thompson, 1939). To the north, Sanders
made limited tests near Chetumal, Quintana Roo, but most of his material
is of later date than the San Estevan finds (Sanders, 1960, pp. 203-8).
In west central British Honduras, there are two significant ceramic
studies in the Belize River valley. The first, of more limited scope, is Thomp-
son's work at Benque Viejo (Thompson, 1940). The other is the analysis
by J. C. GifTord of approximately 200,000 sherds from nearly 70 separate
stratigraphic tests at Barton Ramie (Smith, Willey, and GifTord, 1960;
Willey and GifTord, 1961; Willey, Bullard, Glass and GifTord, 1965).
Although I have been able to examine the type collections and have profited
from discussions with GifTord, the Barton Ramie pottery analysis has
not yet been described, and detailed comparisons with it are consequently
difficult.
The most detailed published study of a large collection is R. E. Smith's
analysis of the Uaxactun pottery (R. E. Smith, 1955). This remains the
prime source of comparative data and is still the "standard" for the Peten
Region. Recent work at Tikal (Culbert, 1963) will doubtless refine the
Uaxactun sequence, but not alter its essentials. Other studies of subordinate
value to the present work include Thompson's of Mountain Cow, British
46
Honduras (1931), and Vaillant's of Holmul in Peten (Mcrwin and
Vaillant, 1932).
Lying to the west and northwest of San Estevan are the large and
elaborate ruins of interior Quintana Roo and Campeche (Ruppert and
Denison, 1943). These are ruins which are possibly of great significance
for understanding the archaeology of northern British Honduras and they
may well hold the key to many of the major chronological and cultural
problems concerning the ancient Maya. But they remain today untouched
by the archaeologist's spade and their ceramic histories are unknown. For
Yucatan, we have the results of Brainerd's surveys (Brainerd, 1958)
supplemented by Sanders' work in Quintana Roo (1960). The Dzibilchal-
tun ceramic sequence and R. E. Smith's study of Yucatan pottery are not
yet available.
THE VASQUEZ COMPLEX
The general charactristics of the Vasquez complex are those of other
Formative stage pottery complexes in the Maya lowlands: monochrome
"waxy" ware, predominance of comparatively simple plate forms, low
necks on storage jars, and decoration by incision in preference to painting.
With respect to colour and other characteristics of the slip, direct compari-
son of sherds reveals that the Vasquez slipped ware has a much stronger
resemblance to the "waxy ware" of the Mamom and Chicanel phases at
Uaxactun than it does to the slipped ware of contemporary phases at
Barton Ramie. Barton Ramie Formative wares tend to be thinner and to
have a somewhat more yellow hue with less contrast between the slip and
the yellowish or buffy undersurface.
When the presence of a Formative pottery complex at San Estevan
was first recognized, we assumed that it would be the temporal equivalent
of such Late Formative complexes as Chicanel and San Jose I. Strati-
graphically, it lay directly beneath a clearly Early Classic complex, the
Barklog. Moreover, nothing in the excavations suggested either an
appreciable gap in time between the two or the existence of an intervening
pottery complex. But placement of the Vasquez complex proved less easy
when detailed comparisons were attempted. While some elements indeed
corresponded to Chicanel and San Jose I, others seemed to resemble more
closely the earlier Mamom complex of Uaxactun.
The Vasquez complex shares with Chicanel and San Jose I, as well
as other Late Formative complexes, such specific traits as a general ten-
dency towards thickened rims, the triangular rim bolster, the composite
silhouette bowl with medial angle, and the simple painted decoration on
some dichrome plates. However, these traits by themselves cannot be
considered diagnostic. The most striking thing in comparing with Chicanel
and San Jose I is the complete absence of such widespread and characteris-
tic Late Formative traits as labial and lateral flanges and vessel supports.
Moreover, widely everted rims are scarce and those that do occur are not
of the deeply grooved horizontal type common at Uaxactun. The absence
of these traits gives an early look to the Vasquez complex and, were
47
San Estevan further to the south, we should suggest that it be placed
coeval with Mamom or at least with very early Chicanel. However, the
absences may be due to geography instead of chronology. Judging by the
large number of presumed Late Formative sherds from Yucatan illustrated
by Brainerd, labial and lateral flanges are scarce in Yucatan. Moreover,
the widely everted rims of Yucatan are mostly of the simple downturned
type which occurs at San Estevan but apparently not at either Uaxactun
or San Jose. Thus, in regard to vessel shape, the Vasquez complex may be
affiliated more with the now poorly known north than it is with the southern
lowlands.
Comparing Vasquez with Mamom, we note that in both plates pre-
dominate over dishes and the Vasquez outcurving-sided plates more
closely resemble their Mamom equivalent than they do the plate forms of
Chicanel. Chamfering, a type of surface treatment restricted to Mamom
at Uaxactun, occurs in Vasquez, although only to modify the basal angles
of plates in a manner not reported at Uaxactun. But Vasquez lacks Mars
Orange, daub ware, figurines and cuspidors (unless some of the "bowls
with vertical or near-vertical sides and thickened rims are of this form).
These are traits which are not only diagnostic of Mamom, but which are
also common in the contemporary Jenney Creek complex at Barton Ramie.
A dichrome in which separate areas of the vessel surface are coloured
red, and cream is fairly common in theVasquez complex. At Uaxactun,
a very similar dichrome is considered diagnostic of Mamom. The tech-
niques and colour ranges are similar, but whereas at Uaxactun it is the
interiors which are most commonly slipped the cream colour, at San
Estevan it is almost exclusively the exteriors which are so coloured. Never-
theless, some individual sherds would not seem out of place at either
site. Dichrome apparently was not found in San Jose 1, which is equiva-
lent in time to Chicanel, unless the Buff Ware of that period be somehow
construed as similar. These indications of early placement do not hold at
Barton Ramie, for there a related dichrome persists into phases contempo-
rary with Chicanel. Moreover, recent work at Tikal may extend the dura-
tion of dichrome in central Peten (Culbert, 1963).
Although the Formative period is often regarded as a period of wide-
spread cultural similarity, regional and local variations in culture were
present during this time and are reflected in the different pottery com-
plexes. For example, San Jose 1, while related to Chicanel, nevertheless
has its own emphases and its own seemingly unique ceramic forms. It is
no surprise to find the same true of Vasquez. Thus, the unslipped jars with
low thick flat-lipped necks have not been reported from elsewhere and seem
to be a purely local trait. Nor should we expect the pottery types at San
Estevan necessarily to represent the same combinations of attributes as do
related pottery types elsewhere. For example, attributes which may be
sorted out in time in central Peten or in the Belize valley may occur
together at San Estevan.
The conclusion which I favour is that Vasquez is more or less contem-
porary with Chicanel and that it represents a conservative Late Formative
pottery complex somewhat similar to the Late Formative complexes of
48
Yucatan. The early look of the complex arises from the retention of such
presumably earlier traits as the dichrome and the simple plate forms and,
especially from the failure to adopt specialized late Formative traits of
more southern distribution such as the labial and lateral flange and the
wide-everted and grooved rim.
An alternative conclusion is that Vasquez actually is contemporary with
late Mamom or very early Chicanel and that a gap exists in the San
Estevan sequence. A third, which I think the least likely, is that the fill
deposits which were excavated were built up of refuse derived from
several Formative periods and that Vasquez represents material contempo-
rary with both Mamom and Chicanel but with the former predominant.
More stratigraphic excavations in northern British Honduras are needed
to confirm the correct placement.
THE BARKLOG COMPLEX
As is generally true in the Maya lowlands, the ceramic break between the
Formative and Classic periods is very marked at San Estevan. The new
slipped ware of the Barklog complex, related to the gloss wares of the
Maya Classic period, is quite different from its "waxy" predecessor and
there is no overlap in the inventory of forms.
The Barklog complex is clearly related to such other Early Classic
manifestations as Tzakol at Uaxactun. San Jose II, and the Hermitage
Phase at Barton Ramie. Diagnostic are the red-and-black-on-orange poly-
chrome basal flange bowls. Other characteristic Early Classic traits include
Z-angle bowls, flaring-sided bowls or dishes with plain rims, scutate
covers for bowls or dishes, slipped jars with vertical or slightly flaring
medium high necks, vessels with inner-inverted feet, and the striated
storage jars with high and medium-high necks.
Particularly interesting is the relationship between the Z-angle and
basal-flange bowls at San Estevan. The latter appear in the Barklog com-
plex only as polychrome and one or two monochrome sherds and, because
of the ware characteristics, all sherds of this form are believed imported.
The local equivalent was the sharp Z-angle bowl, one of the most
abundant forms of the Barklog complex. In the succeeding Trial Farm
complex the picture is different. Then locally-made basal-flange bowls
become a dominant form and replace the sharp Z-angle vessels com-
pletely. Thus the priority of the Z-angle bowl over the basal-flange bowl
is clear at San Estevan. The latter, although made elsewhere and traded
to San Estevan during Barklog times, did not become established as
an indigenous type until later. The basal-flange evidently was introduced
to San Estevan from the same source that provided the intrusive poly-
chromes of the Barklog complex, probably regions to the south or
southwest.
The Z-angle to basal-flange succession is interesting not only for
the light which it throws on the spread of the basal flange form, but also
because it implies a position early in the Early Classic period for the
Barklog complex. At Uaxactun, Z-angle bowls are especially characteristic
49
and outnumber basal flanges in the Tzakol 1 subphase, although they are
reported as a minor form through Tzakol 3. In Yucatan, the form also
seems to belong very early in the Early Classic period to judge by a
collection of Z-angle sherds with designs like Tzakol 1 from near Merida
(Brainerd, 1958, pp. 238-9). Another indication of early position is the
occurrence on Z-angle vessels of mammiform tetrapod feet, a feature of
the Holmul 1 or Matzanel style which falls chronologically between the
Formative and Classic periods (for example, Thompson, 1931, PI. 44,
upper left; Merwin and Vaillant, 1932, pi. 18, c). Such feet never, or
very rarely, occur on basal-flange vessels although they are sometimes
associated with basal-flange vessels in tombs. Although the sharp Z-angle
bowl denotes an early position where it occurs, its distribution is sporadic.
None are reported from San Jose nor from Barton Ramie. Indeed, when
we have better geographical coverage of pottery sequences, the temporal
and spatial relationships between the Z-angle and the basal flange may
throw important light on the beginnings of the Maya Classic period.
Fragments of two polychrome vessels with mammiform tetrapod feet
were found in the Barklog complex. As mentioned, these are features of the
Holmul 1 or Matzanel style. A potstand, also belonging with the complex,
might be considered a feature of this style also. The Holmul 1 style is
most abundantly represented in British Honduras and eastern Peten, and
is believed to be somehow connected with the cultural influences which
triggered the Classic development in the lowland area. Vessels representing
this style have been found in association both with Late Formative deposits
and, more frequently, with Early Classic deposits. Most finds have been in
grave lots. But at Barton Ramie, evidence from house ruin tests demon-
strated a complex of domestic as well as funerary types which were asso-
ciated with the Holmul 1 style. This complex, the Floral Park, is believed
to represent an intrusion of foreign peoples who strongly influenced, but
did not completely break, the indigenous cultural continum from Late
Formative to Early Classic (Willey and Gifford, 1961). In central Peten,
recent work at Tikal shows strong Holmul 1 influence in the Late Forma-
tive "Cimi" complex, but continuities with earlier Formative complexes
are thought too strong to denote introduction of the features by a large-
scale population movement (Culbert, 1963).
Holmul 1 vessels are fairly common in northern British Honduras
(see Introduction of this report). The best documented collection comes
from tombs near Douglas about 10 miles from San Estevan (Anderson
and Cook, 1944). Here Holmul 1 vessels apparently were associated with
Tzakol 1 or 2 basal-flange bowls in tomb assemblages which closely
resembled that of the upper tomb at Holmul.
At San Estevan, we found no evidence for a Holmul 1 "period"
comparable to the Floral Park phase at Barton Ramie. The only traces of
the style were the aforementioned polychrome vessels and possibly the
potstand. The polychrome decoration of the vessels seems somewhat more
developed than that represented on the Douglas specimens, and the vessel
surfaces are more perfectly smoothed and lustrous than on specimens
from Holmul and Barton Ramie which I have handled. At present, we
50
have inadequate evidence concerning the maximum duration of Holmul 1
features into the Early Classic period, but their presence supports a
comparatively early position for the Barklog complex.
In sum, the evidence indicates a position for Barklog in the early
part of the Early Classic period, contemporary with Tzakol 1 and
probably at least part or Tzakol 2.
THE TRIAL FARM COMPLEX
The Trial Farm complex cannot be fixed in time with the same accuracy
as the Barklog complex. Most of the material comes from stratigraphically
terminal deposits, including units in the present topsoil zone, and the
terminal date of the complex is necessarily open. In surface and other
deposits created mainly by erosion and collapse after abandonment,
types of various periods may be mingled together. Moreover, much of the
material is very weathered and fragmentary, making knowledge of ware,
decoration, and complete vessel forms particularly difficult.
The slipped and unslipped ware of the Trial Farm complex seems
to have been unchanged from that of Barklog. Continuity is also suggested
by the necks of the unslipped jars. But nearly all other forms are new.
The characteristic Trial Farm vessel form, found in all provenience
units, is the basal-flange bowl. It is manufactured with the same paste
as the Z-angle bowls of Barklog, but, significantly, no sherd of the latter
was found in a Trial Farm provenience. The basal-flange bowl is a hall-
mark of the Early Classic period in the Maya area. It implies contempora-
neity of Trial Farm with such phases as Tzakol, San Jose II, Holmul II to
IV, and Hermitage at Barton Ramie. Since there is good evidence for
placing Barklog in the early part of the Early Classic, Trial Farm should
fall in the later part and be contemporary with Tzakol 3. Other vessel
forms which conform to this position are the barrel-shaped vases and out-
curving-sided plates found in Burial 1 and Cache 2. These are forms
frequently associated with Early Classic caches and burials.
But other Trial Farm forms are more at home in the Late Classic
period. The bowls or dishes with incurving sides resemble San Jose III
to V and Tepeu forms more than they resemble any San Jose II or Tzakol
form. The simple silhouette flaring-sided bowls, such as the example illus-
trated in Figure 16, g, is a Period III form at San Jose (Early Late Classic),
although at Uaxactun it seems to be Tzakol 3. At San Estevan, the most
complete example which was found was associated with a ridged plate
(Fig. 16, e), which is a marker for San Jose III and Tepeu I, so an early
Late Classic position seems to be best. A definite Late Classic trait at
San Estevan is the mold-made figurine whistle, represented by one battered
but indisputable example. Other forms, such as the bowls or dishes with
bolstered or thickened rims, and the flaring-sided vessels with wide interior
rim bevels suggest Late Classic modes, but the vessel forms are too poorly
known to allow comparison. Still other rim sherds could belong to either
Early or Late Classic forms.
The construction fills of the final building period represented in the
51
excavations were remarkably free of refuse and sherds. Accordingly, the
bulk of the Trial Farm sherds must have come from deposits associated
with the occupation of the final structures. The actual building of Construc-
tions I-A and II-A must have been late Early Classic. All of the pottery
offerings with Burial 1, which is contemporary with the completion of I-A,
fit well into this period. Furthermore, in Structure II the stratigraphic
evidence indicates that II-A was built up at the end of the time of the
Barklog complex, which would be in the middle or later part of the Early
Classic.
The prevalence of basal-flange forms in debris deposits formed after
the abandonment of I-A and II-A suggests that the most intensive occupa-
tion was also late Early Classic. Other forms, such as the incurving-sided
bowls, suggest continuation of occupation into the Late Classic contem-
porary with San Jose III-Tepeu I and perhaps later. In other words, one
suggested interpretation is that the Trial Farm complex is a mixed
assemblage representing several ceramic periods spanning the last part of
the Early Classic and much of the Late Classic.
Another interpretation is that Trial Farm is a transitional complex,
occupying a period of time when both Early Classic and Late Classic
pottery types were in use together. Still another alternative, which I believe
least probable, is that Trial Farm is predominantly a Late Classic complex
and that in northern British Honduras the basal-flange form persisted later
than it did in central Peten and in west-central British Honduras. Such a
survival would accord with Rands' suggestion that the reappearance of
the basal flange in notched or stepped form in Tepeu 2 implies its con-
tinuing elsewhere through the time of Tepeu I (Rands, 1961, p. 335).
Whatever the correct explanation, the surprising thing is the strong,
apparently Early Classic representation in the final pottery complex which
we could discover stratigraphically. The impression is that occupation
tapered off and that there was no new construction during the Late Classic
period. In other parts of British Honduras and Peten where excavations
have been carried out, the Late Classic emerges as a time of population
growth and ceremonial centre expansion, and this situation is reflected in
the great abundance of Late Classic pottery types in the final deposits of
the sites. Whether the history of San Estevan and northern British Honduras
was truly different, or whether the Trial Farm complex results from abnor-
mal conditions of deposition, remains for future excavations to decide.
The suggested chronological correlation of the San Estevan pottery
complexes with phases at Uaxactun and San Jose is summarized in Figure
17.
Stone and Miscellaneous Artifacts
The most lucrative sources in a Maya site for artifacts are tombs and
ceremonial caches. At San Estevan, the single burial found was not a par-
ticularly rich one and of the two caches discovered one contained only a
52
pottery vessel and the other only unworked shell. Thus the artifact yield
was meagre. Utilitarian implements of chipped Hint predominated, followed
by tools and ornaments of ground stone. No worked bone or shell was
encountered. In the descriptions which follow, dimensions given are the
maximum dimensions unless otherwise noted. The numbers of each artifact
type found with each pottery complex are also given.
OBJECTS OF CHIPPED FLINT
The chipped implements are made from a good quality flint which is gray
or tan in colour. Rare pieces are dark brown or milky white. A few pieces
have pinkish or brownish veins.
celts (PI. XVII, upper left and upper centre)
Pointed at one end, rounded or squared at the other. Bifacially worked
with retouching of edges usual. Length (2 complete specimens) : 15.0 and
17.0 cm.; Width (5 specimens): 6.4 to 7.5 cm.; Thickness (8 specimens):
1.8 to 3.5 cm., average 2.2 cm. Of 4 complete blunt ends, 2 show signs of
battering and a third use-polish. Number'. Vasquez, 0; Barklog, 2; Trial
Farm, 5. Points probably from this implement: Vasquez, 3; Barklog, 3;
Trial Farm, 4.
The distribution and purpose of this implement have been discussed
by Kidder (1947, p. 5) and W. Coe (1959, p. 11). They refer to it as a
"general utility tool" or "chopper." Its distribution appears to be restricted
to the Maya lowlands. At least in Yucatan, it goes out of use at the end of
the Classic period or shortly thereafter. In British Honduras and Peten it
is probably the most common utilitarian stone tool to be encountered in
Maya ruins. I have picked many from the ground surface while travelling
on horseback through Peten and once ran across a manufacturing site
where the ground was covered with a litter of celt fragments and spalls.
Kidder believes the polish on some examples is due to protracted use in the
earth presumably as an agricultural tool. Other specimens are believed to
have been used for stone working or chopping because of battering of the
rounded end. It is difficult to avoid the conclusion that they were primarily
forest-clearing and wood-cutting tools. Living as they did in a heavily
forested tropical land, the ancient Maya must have had to clear land and
cut wood quite constantly, and no other tool as generally suitable for the
purpose as the chipped celt has been found. Coe notes that no consistent
pattern of abrasions has been found to demonstrate conclusively that the
celts were hafted. Nevertheless, it seems likely that the pointed butts of the
celts were sometimes inserted through wooden hafts to form a hatchet-like
implement resembling in shape the ceremonial models of hatchets made of
clay, obsidian (Thompson, 1939, p. 171, PI. 28a), and slate (from Barton
Ramie) which have been found in British Honduras.
drills or picks ? (PI. XVII, upper right)
Thinner and more gradually tapering than celts; bifacially worked with
secondary chipping at the edges. No complete specimens were found.
53
Length of longest fragment: 9.6 cm.; Width: 2.0 to 3.4 cm.; Thickness:
1.2 to 1.9 cm. Number: Vasquez, 1; Barklog, 3; Trial Farm, 1.
pecking tools ? (PL XVII, centre row, centre)
Small, rather crudely chipped bifacial tools which are pointed at one end.
One complete specimen has secondary retouch along the edges, and the
point and the edges are somewhat battered. The other specimen has a
similar shape but no retouch. Length: 6.7 cm. and about 8.5 cm. (esti-
mated); Width: 3.2 and 3.5 cm.; Thickness: 1.6 and 2.3 cm. Number:
Barklog, 1 ; Trial Farm, 1 .
three-pointed tool (PI. XVII, centre left)
A rather crudely chipped, but apparently carefully shaped, bifacial tool
with three sharp points. Some retouching of edges. Length: 7.6 cm.; Width:
4.9 cm.; Thickness: 1.9 cm. Number: Vasquez, 1.
tanged point (PL XVII, centre right)
Made from a large flake. The inner surface (in relation to the parent core)
is slightly curved and is unworked except for the tang which is bifacially
worked. The outer surface is convex with secondary working along the
edges. The point is broken from this specimen and the edges are somewhat
battered. Length of fragment, including tang: 15.4 cm., original length
about 19 cm.; Width: 1.1 cm.; Thickness: 1.7 cm. Number: Barklog, 1.
The distribution of this distinctively shaped artifact has been discussed
by Kidder (1947, p. 19) and W. Coe (1957). It is found in the southern
Maya lowlands and has been reported most frequently from British Hon-
duras. It also occurs in lower Central America and, surprisingly, in Haiti.
CHIPPED AND POLISHED POINT
The pointed tip of a probable awl-like implement, chipped and well
polished, possibly from use. Length of fragment: 2.3 cm. Number: Trial
Farm, 1.
hammerstones (PL XVII, lower row)
Of two specimens, one is approximately spherical with diameters 6 to 8 cm.
and is battered on nearly the entire surface. The other is more-or-less
discoidal with one battered side, diameter 9.5 cm.; thickness, 4 cm.
Number: Trial Farm, 2.
Hammerstones of the spherical type are reported from Piedras Negras
(Coe, 1959, p. 12) but not from Uaxactun or San Jose.
PLANO-CONVEX SCRAPERS
A group of 5 artifacts, at least 3 of which are fragmentary, which share
little in common except that all are made from large flakes, are worked on
only one face, and have one or more edges of the convex surface retouched.
Specimens range from about 3 cm. to about 7 cm. in length. Number:
Vasquez, 1 ; Barklog, 3; Trial Farm, 1.
WORKED FLAKES
Irregular flakes showing use chipping on one or more edges were found
sparingly in deposits of all periods.
54
OBJECTS OF OBSIDIAN
The only obsidian artifacts found at San Estevan were flake blades. These
were the common small cutting tools of the ancient Maya. The material
was imported mainly from the Guatemalan highlands. At San Estevan, all
are of black obsidian, quite transparent when thin, and several specimens
have light black streaking. All specimens are fragments whose edges are
nicked by use. Width: .05 to .14 cm., average .09 cm. Number: Vasquez,
4; Barklog, 10; Trial Farm, 5.
OBJECTS OF GROUND STONE
METATES (Fig. 18, a, b)
Two small fragments from probably leg-less metates. One is of pink
granite and apparently was turtle-backed in shape with a concave grinding
surface (Trial Farm; Fig. 18, b) . The other is of a fine-grained gray-green
granitic stone and had a concave grinding surface, a flat bottom and rather
irregularly finished sides (Vasquez; Fig. 18, a). A fragment of vesicular
basalt found on the ground surface may also be from a metate. Number:
Vasquez, 1; Trial Farm, 1.
manos (Fig. 18, c, d)
Two fragments of limestone manos. One, from a central portion of the
implement, is plano-convex in cross-section with three surfaces polished
by use. Width: 1 A cm.; Thickness: 5.3 cm. (Trial Farm; Fig. 18, c). The
other is wedge shaped in cross-section with wear on two surfaces, end is
blunt. Width: 6.1 cm.; Thickness: 3.5 cm. (Trial Farm; Fig. 18, d) .
One pink granite fragment is probably from a mano with plano-convex
cross-section with one surface polished from use, thickness about 4.5 cm.
(Vasquez). The pink granite of which this mano and one of the metate
fragments are made comes from the Mountain Pine Ridge in the Cayo
District of British Honduras.
BARK BEATER (Fig. 18, e)
The single specimen is of limestone and about one-half complete. There is
a hafting groove around the rim. One surface is scored, the other is plain.
Width: 7.5 cm.; Thickness: 3.9 cm.; Number: surface find, 1.
ear plug (Fig. 18, g)
Fragment of an ear-plug flare of a fine grained gray stone. The interior is
grooved and the entire object highly polished. Diameter: 3.6 cm.; Width:
2.0 cm. (Trial Farm).
JADE PENDANT (Fig. 18, /)
A simple jade pendant more highly polished on one surface than the other.
A bi-conical suspension hole is drilled transversely through the top.
Length: 2.7 cm.; Width: 1.6 cm.; Thickness: 0.6 cm. Found with Burial 1
(Trial Farm).
55
DENTAL INLAYS
Jade chips 2.0 to 2.5 mm. in diameter and 1.0 to 1.5 mm. thick set into
the medial and lateral incisors, canines and first pre-molars of the upper
jaw of Burial 1 (Trial Farm) .
WORKED SHERDS
PERFORATED DISK
Irregularly circular, diameter 3.5-4.0 cm. In the centre a bi-conical per-
foration. Number: Vasquez, 1.
IMPERFORATED DISKS
Of three specimens, one had a diameter of only 2.0 cm. (Barklog). The
other two had diameters of 3.5 to 4.0 cm. and were found together (Trial
Farm).
Similar perforated and unperforated disks made from potsherds are
reported from San Jose (Thompson, 1939, p. 153, Fig. 91, n, r), Benque
Viejo (Thompson, 1940, Fig. 55), Uaxactun (Ricketson and Ricketson,
1937, p. 218; Kidder, 1947, p. 68), Piedras Negras (W. Coe, 1959, p. 69),
and Copan (Longyear, 1952, pp. 103-4), and are probably common every-
where. The perforated ones may have been spindle whorls, the unperforated
ones gaming pieces, unfinished spindle whorls, or — when large enough —
lids for small-mouthed containers.
MATTING (Fig. 6)
In Construction I-C, woven mats had been placed over the roof beams
before laying of the plaster capping of the roof. Fallen fragments of plaster
retained the impressions of the mats, which had themselves decayed. The
weave was a simple over-two and under-two twill, the individual cane
elements being approximately 1 cm. wide. Similar evidence was also found
of mats of the same weave but with the elements about 5 cm. wide and
each element composed of 6 strands.
Summary and Conclusions
SUMMARY OF THE FINDINGS
During 1962, work was carried out at a medium sized ancient Maya
ceremonial centre at San Estevan, on the New River in northern British
Honduras. Two adjacent mounds, Structures I and II, were excavated
stratigraphically in order to determine the cultural sequence. Remains were
found spanning the range in time from the Late Formative period to the
Late Classic period (from probably before the time of Christ until the
eighth or ninth centuries A.D.).
Three building periods were found in Structure I. These were designated,
from early to late, Constructions I-C, I-B, and I-A. The latest, Construc-
56
tion I-A, was almost completely excavated; the others only partly. Con-
struction I-C is believed to date from the very beginning of the Early
Classic period. Constructions I-B and I-A appear to have been built later
in the Early Classic period. No construction dating from the Late Classic
period was identified in either of the two mounds excavated.
Construction I-C was probably "T'-shaped in its entirety, but the main
portion was a small masonry-walled building, probably a temple, standing
on a substructure platform not quite 2 m. high. The building had two
rooms, windows, and an interior staircase which formerly led to the top of
a beam-and-mortar roof. The substructure had apron and basal mouldings
and the entire structure had been painted red.
Construction I-B was only trenched. It was larger and higher than I-C,
which it covered, and it may have had three superstructure rooms, one
behind the other. The latest component of Structure I, Construction I-A,
had a substructure platform more than 4 m. high with a broad stairway
across the front. The superstructure had two rooms, one behind the other,
with the back room at a higher level than the front room. The latter was
15 m. long and had three doorways opening onto the stairway; the back
room was smaller and was entered by one doorway. I-A probably had a
beam-and-mortar roof. A burial, the only one found at San Estevan, was
associated with Construction I-A.
In Structure II, two building periods were identified. Associated sherds
indicated that Construction II-B, the earliest, was built at the same time as
the earliest construction in Structure 1, that is, at the beginning of the early
Classic period. Too little of II-B was cleared to explain its plan, but it
consisted of a superstructure building on a substructure about 3 m. high
and, like I-C, had been painted red. The superimposed Construction II-A
was built later in the Early Classic period and was in use contemporary
with the latest construction of Structure I, Construction I-A. II-A had a
higher substructure than II-B and a superstructure composed of a long
front room and two or three back rooms placed side by side at a higher level
than the front room. Like I-A, it probably had a beam-and-mortar roof.
The architectural stratigraphy at San Estevan was clear cut and three
distinct pottery complexes were associated with the building remains. From
early to late, the pottery complexes are (1) Vasquez, (2) Barklog, and
(3) Trial Farm.
In its general characteristics, the Vasquez complex resembles Formative
period pottery complexes in the Maya area. It features monochrome "waxy"
ware, predominance of comparatively simple plate forms, low necks on
storage jars, and decoration by incision in preference to painting. The
complex is thought to fall into the Late Formative period and to be approxi-
mately contemporary with the Chicanel phase at Uaxactun and with San
Jose I. However, certain widespread and characteristic Late Formative
pottery features are lacking, for example, labial and lateral flanges and
vessel supports. Moreover, wide everted rims are comparatively scarce.
These absences, plus the prevalence of a red and cream dichrome, give the
Vasquez complex an early look and suggest the possibility of contem-
poraneity with Mamon or early Chicanel. But the interpretation is favoured
57
that geography, rather than chronology, is responsible for the absences
since the same features apparently are absent or scarce in Late Formative
pottery of Yucatan. The Vasquez complex may relate more closely with
that area than to the more southern parts of the Maya area, but it is
acknowledged that more work is needed in northern British Honduras to
confirm the relationships.
The Barklog complex is sharply differentiated from the Vasquez com-
plex in ware characteristics, vessel forms, and decoration. It belongs in the
early part of the Early Classic period contemporary with Tzakol I to II at
Uaxactun. The outstanding vessel form is the sharp Z-angle bowl. Others
include simple flaring-sided bowls or dishes, scutate covers, vessels with
inner-inverted feet, and slipped jars with medium high cylindrical necks.
Sherds of polychrome basal-flange bowls are also present, but are believed
to be imported because of ware characteristics. No basal-flange bowls
made with the local paste occur in the Barklog complex, their place being
taken by the Z-angle bowl. Holmul 1 or Matzanel traits present in the
Barklog complex include two mammiform tetrapod vessels and possibly a
potstand, but no evidence was found for a distinct complex of such material
at San Estevan.
The Trial Farm complex comes from deposits which are stratigraphi-
cally terminal and which are believed to have been derived mainly from
the occupation, rather than the construction fills, of the excavated mounds.
The sherds are very badly weathered. The outstanding form in all deposits
is the basal-flange bowl which, in contrast with the basal-flange bowls of
the earlier Barklog complex, is now made in the local paste. The basal-
flange bowl is a hallmark of the Early Classic period, but other common
forms, such as bowls or dishes with incurving sides, fit best in a Late Classic
context. Trial Farm probably spans a considerable length of time and,
except as a stratigraphic division at San Estevan, its status as a meaningful
pottery "complex" is dubious. Its main importance is to indicate that the
final construction in this part of the San Estevan site was probably in the
later part of the Early Classic period and that occupation and use of the
buildings diminished in intensity thereafter.
Non-ceramic artifacts were comparatively few and unspectacular. We
found no rich caches, and the only burial was not a rich one. No worked
bone or shell was found. Most common were simple utilitarian tools of
chipped flint, especially the chipped celt which was probably the common
forest clearing tool of the lowlands. Obsidian flake blades and pieces of
metates and manos, including examples made of Mountain Pine Ridge
granite, turned up. The one burial had jade inlays in the teeth and a small
jade pendant, but luxury items were otherwise practically absent.
CONCLUDING REMARKS
Our work has demonstrated that the period of occupation of the San
Estevan ceremonial centre was several centuries long. Moreover, since
we did not plumb the deepest deposits at the site, it is possible that there
are remains even earlier than the presumably Late Formative Vasquez
58
complex. It is no surprise to find that the earliest occupation was during
the Formative period. Remains of that period have been found underlying
all large sites which have been excavated in the southern Maya lowlands.
President evidence is that the Maya population was spreading through the
Yucatan peninsula in the Middle Formative period (ca. 700-300 B.C.) and
that during Late Formative times (ca. 300 b.c.-a.d. 300) they were fully
established in all the regions and localities where we find them during the
Classic period. Evidently northern British Honduras was no exception.
The ceramic changes which take place between the Vasquez and
Barklog complexes correspond in general to the changes which occur in
other parts of the lowland Maya area between the Late Formative and
Early Classic periods, but one cannot fail to be impressed at the complete-
ness and abruptness of the changes. In effect, one long-lived ceramic
tradition comes to an end and another begins. It has been felt that the
Holmul 1 , or Matzanel, style has something to do with these changes and
a separate period, the Protoclassic, is sometimes set aside for it. At San
Estevan, Holmul 1 attributes turned up on two vessels of the Barklog
complex, but we found no evidence for a distinctive associated pottery
complex which was comparable to the Protoclassic Floral Park phase at
Barton Ramie (Willey and Gifford, 1961). If a population intrusion was
responsible for bringing the Holmul 1 style to the lowland Maya area, then
San Estevan, like Uaxactun, was evidently outside the area of intrusion and
must have received its influences indirectly.
While the Vasquez-Barklog transition represents a break in tradition,
the Barklog-Trial Farm transition represents continuity. Especially interest-
ing here is the evidence that San Estevan adopted the ubiquitous basal-
flange form as a local product only comparatively late in the Early Classic
period.
The most surprising result of the San Estevan work was the feeble
representation of the Late Classic period. Not only could we not subdivide
the cultural material from the Late Classic period, but we could not even
obtain a good stratigraphic separation between Early and Late Classic.
Although other explanations are possible, we have favoured the interpreta-
tion that the final construction was during the late part of the Early Classic
period and that use of the buildings diminished from then until abandon-
ment sometime during the Late Classic. This contrasts with the picture from
excavated sites in central British Honduras and Peten. There, the Late
Classic seems to have been the time of greatest population, and pottery and
other artifacts of that period are the most abundant in the upper levels of
the sites. Did the Classic Maya decline begin in northern British Honduras
relatively earlier than in other parts of the southern lowlands? Or is the
situation the result of purely local factors — perhaps even restricted to part
of the San Estevan site?
We found no objects assignable to the post-Classic period, although
San Estevan is in the midst of an area where numerous late post-Classic
incensarios and other items have been found. It is also worthy of note that
among the numerous published illustrations of antiquities collected from
northern British Honduras there appear to be no objects of obviously early
59
post-Classic origin. What seems the most likely is that northern British
Honduras, like most of the southern lowlands, was abandoned during or at
the close of the Late Classic period and the post-Classic finds represent a
comparatively late movement back into the area, probably from the north
or west, of the Maya who formed the populace of the native province of
Chetumal.
The San Estevan excavations produced a poor representation of the finer
products of ancient Maya civilization, and it is hard to avoid the impression
that the site occupied a relatively marginal and provincial status. But the
results of one season of test excavations provide only a part of the evidence
for reconstructing the culture history of northern British Honduras. We will
need more knowledge concerning the region than we now have before our
findings can be placed in their proper perspective.
REFERENCES
ANDERSON, A. H., AND H. J. COOK, 1944
Archaeological Finds near Douglas, British Honduras, Carnegie
Institution of Washington, Notes on Middle American Archaeology
and Ethnology, Vol. 2, pp. 83-92, Cambridge.
BORHEGYI, S. F., 1959
"The Composite or 'Assemble-it-yourself Censer: A New Lowland
Maya Variety of the Three-Pronged Incense Burner," American
Antiquity, Vol. 25, No. 1, pp. 51-8, Salt Lake City.
BRAINERD, G. W., 1958
The Archaeological Ceramics of Yucatan (Univ. of California,
Anthropological Records, No. 19), Berkeley and Los Angeles.
BULLARD, W. R., 1960
"Maya Settlement Pattern in Northeastern Peten, Guatemala,"
American Antiquity, Vol. 25, No. 3, pp. 355-72, Salt Lake City.
1961
"Archaeological Investigation of the Maya Ruin of Topoxte, Peten,
Guatemala," American Philosophical Society, Year Book 1960,
pp. 551-4, Philadelphia.
BURDON, J. a., 1931
The Archives of British Honduras, 3 Vols., London.
CASTELLS, F. DE P., 1904
"The Ruins of Indian Church in British Honduras," American
Antiquarian and Oriental Journal, Vol. 26, No. 1, pp. 32-7,
Chicago.
CHAMBERLAIN, R. S., 1948
The Conquest and Colonization of Yucatan, J 5 17-1550 (Carnegie
Institution of Washington, Publication 582), Washington.
coe, w. r., 1957
"A Distinctive Artifact Common to Haiti and Central America,"
American Antiquity, Vol. 22, No. 3, pp. 280-2, Salt Lake City.
60
1959
Piedras Negras Archaeology: Artifacts, Caches, and Burials (Uni-
versity Museum, University of Pennsylvania, Museum Monographs),
Philadelphia.
CULBERT, T. P., 1963
"Ceramic Research at Tikal, Guatemala," Ceramica de Cultura
Maya, Vol. 1, Nos. 2-3, pp. 34-42, Cambridge, Mass.
ESCALONA RAMOS, A., 1946
"Algunas Ruinas Prehispanicos en Quintana Roo," Boletin de la
Sociedad Mexicana de Geografia y Estadistica, Vol. 61, No. 3,
pp. 513-628, Mexico.
GANN, t. w. f., 1895-97
"The Contents of Some Ancient Mounds in Central America," Pro-
ceedings of the Society of Antiquaries of London, 2nd ser., Vol. 16,
pp. 308-17, London.
1901
"Mounds in Northern Honduras," Bureau of American Ethnology,
19th Annual Report, Part 2, pp. 655-92, Washington.
1911
Exploration Carried on in British Honduras During 1908-9 (Uni-
versity of Liverpool, Annals of Archaeology and Anthropology,
Vol. 4, Nos. 2-3), pp. 72-87, Liverpool.
1918
"The Maya Indians of Southern Yucatan and Northern British
Honduras," Bureau of American Ethnology, Bulletin 64, Wash-
ington.
1926
Ancient Cities and Modern Tribes: Explorations and Adventures in
Maya Lands, Scribners, New York.
1927
Maya Cities: A Record of Exploration and Adventure in Middle
America, Duckworth, London.
GANN, T. W. F., AND M. GANN, 1939
"Archaeological Investigations in the Corozal District of British
Honduras," Bureau of American Ethnology, Bulletin 123, Anthro-
pological Paper, No. 7, Washington.
HABERLAND, W., 1958
"An Early Mound at Luisville, British Honduras," Man, Vol. 58,
pp. 128-9, London.
KIDDER, a. v., 1947
The Artifacts of Uaxactun (Carnegie Institution of Washington,
Publication 576), Washington.
longyear, j. m. in, 1952
Copan Ceramics: A Study of Southeastern Maya Pottery (Carnegie
Institution of Washington, Publication 597), Washington.
61
LOPEZ COGOLLUDO, D., 1867-68
Historia de Yucatan, 3rd Edition, Merida (1st edition, Madrid,
1688).
MERWIN, R. E., AND G. C. VAILLANT, 1932
The Ruins of Holmul, Guatemala {Peabody Museum, Harvard
University, Memoirs, Vol. 3, No. 2), Cambridge.
POLLOCK, H. E. D., R. L. ROYS, T. PROSKOURIAKOFF, AND A. L. SMITH, 1962
Mayapan, Yucatan, Mexico (Carnegie Institution of Washington,
Publication 619), Washington.
rands, R. l., 1961
"Elaboration and Invention in Ceramic Traditions," American
Antiquity, Vol. 26, No. 3, pp. 331-40, Salt Lake City.
RICKETSON, O. G., AND E. B. RICKETSON, 1937
Uaxactun, Guatemala: Group E — 1926-1931 (Carnegie Institution
of Washington, Publication 477), Washington.
ROYS, r. l., 1957
The Political Geography of the Yucatan Maya. (Carnegie Institu-
tion of Washington, Publication 613), Washington.
RUPPERT, K., AND J. H. DENISON, 1943
Archaeological Reconnaissance in Campeche, Quintana Roo, and
Peten (Carnegie Institution of Washington, Publication 543),
Washington.
SALISBURY, S., JR., 1879
"Terra Cotta Figures from Isla Mujeres," in The Mexican Calendar
Stone and other Papers, compiled and arranged by Stephan Salis-
bury, Jr., Worcester, Mass.
SANDERS, W. T., 1960
Prehistoric Ceramics and Settlement Patterns in Quintana Roo,
Mexico (Contributions to American Anthropology and History, No.
60, Carnegie Institution of Washington, Publication 606), Washing-
ton.
SCHOLES, F. V., C. R. MENENDEZ, J. I. RUBIO MANE, AND E. B. ADAMS (EDS.),
1936-38
Documentos Para La Historia de Yucatan, 3 parts, Merida.
SHEPARD, A. O., 1961
Ceramics for the Archaeologist (Carnegie Institution of Washing-
ton, Publication 609), Washington.
SMITH, r. e., 1955
Ceramic Sequence at Uaxactun, Guatemala (Middle American Re-
search Institute, Tulane Univ., Publication No. 20), New Orleans,
search Institute, Tulane University, Publication No. 20), New
Orleans.
SMITH, R. E., G. R. WILLEY, AND J. C. GIFFORD, 1960
"The Type-Variety Concept as a Basis for the Analysis of Maya
Pottery," American Antiquity, Vol. 25, No. 3, pp. 330-40, Salt
Lake City.
62
THOMPSON, J. E. S., 1931
Archaeological Investigations in the Southern Cayo District, British
Honduras (Field Museum of Natural History, Anthropological
Series, Vol. 17, No. 3), Chicago.
1939
Excavations at San Jose, British Honduras (Carnegie Institution of
Washington, Publication 506), Washington.
1940
Late Ceramic Horizons at Benque Vie jo, British Honduras (Con-
tributions to American Anthropology and History, No. 35, Carnegie
Institution of Washington, Publication 528), Washington.
VILLAGUTIERRE SOTOMAYOR, J. DE, 1933
Historia de la Conquista de la Provincia de El Itza, Reduccion, y
Progressos de la de El Lacandon, y Otros Naciones de Jndios
Bdrharos, de la Mediaciones de El Reyno de Guatemala, a las
Provincias de Yucatan, en la America Septentrional, 2nd edition
(Sociedad de Geografia e Historia, Bihlioteca "Goathemala" Vol.
9), Guatemala (1st edition, Madrid, 1701).
WILLEY, G. R., AND J. C. GIFFORD, 1961
"Pottery of the Holmul I Style from Barton Ramie, British Hon-
duras," Essays in Pre-Columbian Art and Archaeology, pp. 152-70,
Harvard University Press, Cambridge.
WILLEY, G. R., W. R. BULLARD, J. B. GLASS, AND J. C. GIFFORD, 1965
Prehistoric Maya Settlements in the Belize Valley (Peabody
Museum, Harvard University, Papers, Vol. 54), Cambridge.
WRIGHT, A. C. S., D. H. ROMNEY, R. H. ARBUCKLE, AND V. E. VIAL, 1959
Land in British Honduras: Report of the British Honduras Land
Use Survey Team (Colonial Research Publications, No. 24), The
Colonial Office, London.
63
j_Dzibilchaltun I
Campeche V' c^^f)
.-■ w i m >
^^/, ■»■ f ! Uoxoctun u^, . IB
J. IMPORTANT ARCHAEOLOGICAL SITES
• MINOR SITES AND TOWNS
NTERNATIONAL FRONTIER
5 O 5 10 15 20
MILES
5 10 15 20 25 30
KILOMETERS
Fig. 1 Map of Northern and Central British Honduras.
64
F/'g. 2. San Estevan Ruin, Plan and Cross-sections.
65
C/3
66
Fig. 4.
Structure I, Plan and East-West Cross-section (Section b-b') of Construction
C. For Section a-a', see Fig. 3; for Sections c-c\ d-d', e-e' see Fig. 5.
67
CLAY
Elevation
c-c
d-d
SECTIONS
Fig. 5. Structure I-C.
a, Elevation and Cross-section (Section e-e') of Middle of South Wall.
b, Elevation of East Wall and Cross-sections of South Wall (Section c-c') and
East Wall (Section d-d').
FINISHED SURFACE
MAT
IMPRESSIONS
CROSS-SECTION
Fig. 6. Structure I, Mat Impressions from Construction I-C.
a, Weave and typical cross-sections of plaster mat impressions from roof.
b, Six strand matting.
.5
1 I
'"•/
N -*
■7
FLOOR V
\
h-
Fig. 7. Structure I, Substructure Corner of Construction I-B(?)
68
o
Z>
\-
o
o
Fig. 8. Structure I, Plan and East-West Cross-section of Construction I-A. Cross-
section runs through Room 1.
69
g
70
** .~
u o?£ .••" ■
o^;
^
^ /
0.V
CONSTRUCTION A
Mag. N
METERS
, EDGE OFTRENCH,,
TANDING WALL HEIGHTS
3IVEN IN CENTIMETERS
ABOVE FLOOR 4
CONSTRUCTION B
METERS
Wig. 10. Structure II, Plans of Constructions A and B. For Cross-sections, see Fig. 9.
a. Plan of Construction A.
h. Plan of Excavated Part of Construction B.
71
Fig. 11. Vasquez Complex Slipped Forms. (Scale: 1/3)
a, Plates with outcurving sides; b, Dishes with medial angles; c. Hemi-
spherical bowls or dishes; d, Bowls with vertical or near vertical sides and
thickened rims; e, Bowls with slightly restricted orifices; /, Jars with low
necks.
72
17 18 * "
19 20 21 "" 22 23 24
L
26
"S
25
I
27
^
29
>\ \
31
30
K
1 2
TUT T^ *v/
73
Fig. 12. Vasquez Complex Slipped and Unslipped Forms, Barklog Complex Poly-
chrome Form. (Scale: 1/3)
a-j, Vasquez complex; k, Barklog complex.
a, Spout; b, Sherds possibly later than Vasquez; c, Unslipped plates with
flaring sides; d, Unslipped vessels with widely flaring sides; e, Bowls or
dishes with rounded sides; /, Jars with low thick necks; g, Jars with very
low necks or collars; h, Miscellaneous jar sherds; /, Miscellaneous bowls
and jar sherds; /, Adorno; k, Polychrome bowls with tetrapod support.
74
s
vc
JO/ \
U.
9
\ &
75
Fig. 13. Barklog Complex Slipped Forms. (Scale: 1/3)
a, Polychrome basal-flange bowls; b, Polychrome rim sherds; c, Sharp
Z-angle bowls; d, Rounded Z-angle bowls; e, Flaring-sided vessels with
plain rims; /, Bowls with incurving sides and slightly everted rims; g,
Vessels with bolstered or slightly everted rims; /?, Basal-flange bowls; i,
Basal-break bowls with everted rims; /, Covers.
76
'■ TSS
9
11
Fig. 14. Barklog Complex Slipped and Unslipped Forms, Trial Farm Complex
Slipped Forms. (Scale: 1/3)
a-h, Barklog; i, /', Trial Farm, a, Jars with medium-high necks; b, Jar with
finger-indented band on shoulder; c, Potstand; d, Spout; e, Miscellaneous
slipped bowl or vase rims; /, Unslipped striated storage jars; g, Miscella-
neous unslipped jar necks; /;, Vessel with inner-inverted feet; i, Polychrome
and dichrome sherds; /, Basal-flange bowls or dishes.
78
^
\
79
Fig. 15. Trial Farm Complex Forms. (Scale: a, 1/4; b-m, 1/3)
a, Basal-flange bowl with piano-relief decoration; b, Flaring-sided vessels;
c, Bowls or dishes with incurving sides; d, Vessels with bolstered or thickened
rims; e, Bowls with rounded sides and plain rims; /, Cylindrical or slightly
barrel-shaped bowls or vases; g, Barrel-shaped vase or bowl ?; h, Hollow
vessel support; i, Slipped jar necks; /', Unslipped jar necks; k, Strap handle;
/, Bowls with triangular rim bolster; m, Vessels with spikes and pie-crust
80
rims.
\ SCALE
2 3
9 N 10
*
I
V 5 ( 1
C l 6 ' V 7.
T
r
<7
a
{
81
Fig. 16. Trial Farm Complex Vessels from Cache 2, Burial 1, and House Ruin
Outside Main Ruin. (Scale: 1/4)
a, from Cache 2; b-d, from Burial 1; e-g, from House Ruin, a. Barrel-shaped
vase and cover; b, c, Round-sided bowls; d, Basal-flange bowl; e, Dish with
lateral flange; /, Barrel-shaped vase; g, Black-on-red dish (design restored).
82
83
SAN ESTEVAN
UAXACTUN
SAN JOSE
1000
TEPEU 3
S.J. T
800
LATE
CLASSIC
TEPEU 2
s.j. rz
_600
400
TRIAL FARM
TEPEU 1
s.j. in
TZAKOL 3
s.j. n
EARLY
CLASSIC
TZAKOL 2
BARKLOG
TZAKOL 1
^^ "\
>
V MATZANEL
200
A.D.
_0
B.C.
VASQUEZ
CHICANEL
S.J.I
1
is
_200
_400
600
MAN
>
40M
MIDDLE
FORMATIVE
Fig. 17. Chronological Correlation of San Estevan Pottery Complexes.
84
ONE-HALF SCALE
9
FULL SCALE
fig. 18. Ground Stone Artifacts. (Scale: a-e, 1/2; /, g, 1/1 )
a, b, Metate rim sections; c, cl, Mano cross-sections, arrows delimit worn
surfaces; e, Bark-beater fragment; /, Jade pendant; g, Ear-plug fragment.
85
•A *
plate i. Structure XIII from Structure XVI.
.}#%:>
m
plate II. View north from Structure X, showing Structure XIV .
86
iflMF
mmmOKKKm
plate in. Stratigraphy in test through plaza floor. Floors 1I-V show as fine-textured
white hands.
f
PLATE IV. Construction I-C, east side of interior
87
I
plate v. Construction I-C, west side of interior, showing doorway and fallen masonry
pier.
plate vi. Construction I-C, interior side of south wall showing wall features (Lintel
beam of window is a replacement).
88
E vil. Construction I-C. West side of exterior, looking north. Shows substructure
mouldings, doorway through west wall (right), doorway through secondary
wall (top). Most of Stairway 3 has been removed.
89
PLATE
viii. Structure I, floor of Construction l-B (Floor 2a) beneath Construction I- A .
plate ix. Structure I, Construction I-A,
90
91
M \
\ \
\
v
plate xi. Burial 1, upper tier of capstones.
k
PLATE XII.
Burial 1, east end of cist, showing pottery vessels over and near skull.
92
plate xin. Structure II, before excavation. View is from Structure I.
93
- ;*w
xiv. Structure II-B, superstructure features. View looks east.
94
plate xv. Structure 11- A, viewed from Structure I.
PLATE xvi. Structure II- A, Room 2, showing doorway to room 3.
95
A
plate xvii. Chipped stone artifacts (Scale 1/3). Upper row, left to right: Celts
and drill or pick; Centre row: Three-pointed tool, Pecking Tool, Tanged
point; Lower row: hammer stones.
96
7124
tf>
M^
V\
$$81
fill
Bra
LIBRARY
ROYAL ON' "HM
TVNews
OpenLibrary